CN117472007A - Coal bunker and coal mill plug-in depth coal blending control system based on cascade optimization strategy - Google Patents

Abstract

The invention provides a coal bunker and coal mill plug-in deep coal blending control system based on a cascade optimization strategy, and relates to the technical field of energy conservation and environmental protection of power plants. The system comprises a server, a communication module, a primary optimization coal blending module, a secondary optimization coal blending module, a model solving module, a relational database interface and a real-time database interface module which are arranged on the server; the primary coal blending optimization module is used for establishing a primary coal blending mathematical optimization model by taking the coal types and the proportion of each coal bin as optimization variables; the secondary coal blending optimization module establishes a secondary coal blending mathematical optimization model by taking the operation combination and the optimal output of each coal mill as optimization variables on the basis of primary coal blending optimization; the model solving module is used for realizing modeling and solving of a primary coal blending mathematical optimization model and a secondary coal blending mathematical optimization model, solving a coal blending scheme of the coal bin, an optimal combination of the coal mill and a mill operation scheme of an optimal output point, and further realizing control and adjustment of the coal bin and the coal mill.

Description

Deep coal blending control system for coal bunker and coal mill based on cascade optimization strategy

Technical Field

The invention relates to the technical field of energy conservation and environmental protection of power plants, and in particular to a coal bunker and coal mill external deep coal blending control system based on a cascade optimization strategy.

Background Art

At present, my country's power generation system mainly includes a parallel power generation system consisting of thermal power, photovoltaic power, wind power, hydropower and nuclear power. By the end of 2022, thermal power generation accounted for 67%, which is still the main force of my country's power generation and will not change in the short term. Optimizing the energy structure and improving the clean and efficient use of coal are long-term measures to achieve the dual carbon goals. During the construction period, my country's thermal power units were designed for a specific single type of coal. The unit burned this single type of coal with good safety, environmental protection and economy. However, the domestic coal market is complex and changeable, and more and more coal-fired power plants have to purchase some non-designed coal types; in addition, the continued rise in coal prices has also forced power generation companies to purchase some inferior coal types to reduce power generation costs. Therefore, at present, a power generation company's coal yard is often equipped with about 10 types of coal with different coal quality parameters. Based on the changes in power generation load, it is very important to dynamically and reasonably allocate coal to the coal yard every day. Through the reasonable design of the coal quality parameters of mixed coal, the adaptability of the boiler to mixed coal can be broadened, fuel efficiency can be improved, and pollutant emissions can be reduced.

There are generally three ways to blend coal for power plants in my country: 1) Coal yard blending, that is, coal blending is carried out in the power plant's special coal blending yard according to the requirements of load and environmental protection indicators, and then transported to the power plant. This method is conducive to the comprehensive dispatch of power coal and can achieve high-precision coal blending, but the investment and maintenance costs of the coal blending yard are large, and it lacks flexibility and real-time performance for power plants; 2) Coal blending in the coal bunker in front of the furnace, that is, in the coal bunker of the power plant, different types of coal are pre-blended in a certain proportion and then sent to the furnace for combustion. In this method, the number of coal types in a single coal bunker generally determines the number of stackers and conveyor belts. Generally, there are no more than two types of coal in a single bunker. This method is widely used in enterprises; 3) In-furnace blending, or sub-grinding blending or layered blending, that is, different single coals are pulverized in different pulverizers and then sent to different burner layers for in-furnace blending. This is the most convenient way to use, but the various types of coal cannot be evenly mixed, and the combustion stability and environmental protection cannot be effectively guaranteed.

Domestic thermal power companies have generally carried out coal blending work, with the goal of pursuing the optimal coal blending cost under the constraints of ensuring unit safety, environmental protection and driving the planned load of the boiler (power generation load, heating load and steam supply load). Taking coal blending in the form of coal bunkers as an example, according to process equipment, coal blending work should include three stages: coal blending in the coal bunker, coal blending operation in the coal mill and coal blending in the boiler. The coal blending stage of the coal bunker is based on the type and proportion of coal in each coal bunker as the optimization variable. At present, enterprises mainly rely on the expert manual experience model to carry out coal blending in the coal bunker. After specifying the coal bunker coal blending plan and adding bunkers, the coal blending operation stage of the coal mill should be based on the operation combination and output (coal feeding rate) of each coal mill as the optimization variable. At present, enterprises also mainly rely on expert manual experience to determine the combined operation of each mill (stop grinding or cutting grinding), and there is little research on the optimal combined output of the coal mill, which is basically adjusted by the simple function of the output and load of the traditional coal mill. In addition, combined with the burner structure and parameters of the boiler, there is even less research on further optimizing the coal blending process.

At present, the coal blending process of my country's thermal power enterprises generally relies on the manual experience of experts and combines with simple calculation models in Excel tables. The biggest disadvantage of coal blending in coal bunkers that relies on manual experience is that it is difficult to ensure the optimality and speed of the coal blending plan. Professional software using computer optimization technology is urgently needed. In addition, there is a lack of research on cascade optimization methods in theory. At present, whether it is coal blending in coal yards, coal blending in coal bunkers, or coal blending in furnaces, they are theoretically based on single-model single-link optimization technology. There are few studies on multi-link cascade optimization methods for coal bunker preparation, coal mill operation, and combustion in boilers, and there is a lack of multi-link process optimization of coal blending.

The present invention proposes a coal bunker and coal mill external deep coal blending control system based on a cascade optimization strategy. On the basis of the primary optimized coal blending in the coal bunker, the coal mill combination and the output points of each coal mill are adjusted to make the fuel load of the coal mill unit optimally match the planned load demand for secondary optimized coal blending, thereby achieving precise coal blending and achieving the purpose of energy saving and environmental protection.

Summary of the invention

The technical problem to be solved by the present invention is to address the deficiencies of the above-mentioned prior art and to provide a coal bunker and coal mill external deep coal blending control system based on a cascade optimization strategy. On the basis of the primary optimized coal blending in the coal bunker, the coal mill combination and the output points of each coal mill are adjusted to make the fuel load of the coal mill unit optimally match the planned load demand for secondary optimized coal blending, thereby enabling my country's power generation equipment to further develop in the direction of intelligence on the basis of basic automation, and further saving the cost of power generation fuel.

In order to solve the above technical problems, the technical solution adopted by the present invention is: a coal bunker and coal mill external deep coal blending control system based on a cascade optimization strategy, including a server, a communication module, and a primary optimization coal blending module based on a coal bunker, a secondary optimization coal blending module based on a coal mill, a model solving module, a relational database interface module and a real-time database interface module formed by software programming and set on the server;

The relational database interface module reads the corresponding coal quality parameters from the SQL Server relational database storing the coal quality parameter information of each single coal type in the coal yard, and provides the basic coal storage data in the coal yard for the primary optimization of coal blending;

The real-time database interface module reads the corresponding real-time data of operating parameters from the OPENPlant real-time database storing the real-time data of the unit operating parameters, and provides the real-time data basis of the unit for the secondary optimization of coal blending;

The coal bunker-based one-time optimization coal blending module establishes a one-time coal blending mathematical optimization model with the coal types and proportions of each coal bunker as optimization variables during the mixed coal blending stage of the coal bunker; at the same time, the one-time optimization coal blending module transmits the mixed coal types and proportions of each coal bunker to the upper computer of the original reclaimer control system through the communication module, and coordinates the coordinated coal loading of the two reclaimers;

The secondary coal blending optimization module based on the coal mill, based on the primary coal blending optimization, establishes a secondary coal blending mathematical optimization model with the operation combination and optimal output of each coal mill as optimization variables during the coal mixing operation stage of the coal mill; at the same time, the secondary coal blending optimization module transmits the coal mill operation plan of the optimal combination and optimal output point of the coal mill through the communication module, so as to realize the compensation adjustment of the coal feeding rate of the original coal mill control system;

The model solving module, based on the established primary coal blending mathematical optimization model and secondary coal blending mathematical optimization model, realizes the modeling and solving of the primary coal blending mathematical optimization model and the secondary coal blending mathematical optimization model, solves the mixed coal blending scheme of the coal bunker, the optimal combination of coal mills and the mill operation scheme of the optimal output point, and then realizes the control and adjustment of the coal bunker and the coal mill.

Preferably, the coal bunker-based primary coal blending optimization module takes the mixed coal combustion meeting the calorific value, sulfur, volatile matter and ash content indicators as constraints, and takes the economic index of the minimum mixed coal as the objective function to establish a primary coal blending mathematical optimization model, as shown in the following formula:

In the formula, minf(x) is the minimum cost of mixed coal, _cj is the cost of the j-th single coal, that is, the unit price of standard coal, _xj is the blending ratio of the j-th single coal, and n is the total number of coal types in the mixed coal;

According to the actual situation of the enterprise's equipment and coal types, some coal quality parameter characteristic indicators of mixed coal are selected as constraints; assuming that the enterprise has restrictions on the actual calorific value, actual sulfur content, actual ash content and actual volatile matter of the mixed coal entering the furnace, the constraints of mixed coal are:

In the formula, _hj , _sj , _aj , and _vj represent the calorific value, sulfur content, ash content, and volatile matter of each type of coal respectively; H, S, A, and V represent the upper or lower limits of the calorific value, sulfur content, ash content, and volatile matter of the mixed coal scheme respectively; according to the actual situation of the enterprise, coal quality characteristic indicators and other special constraints of the enterprise's own characteristics and needs can also be added to the constraints.

Preferably, the secondary coal blending optimization module based on coal mill takes the real-time fuel cost as the target, introduces the operating parameter variables of each coal mill in combination with the actual situation, and minimizes the real-time fuel cost of power production under the premise of meeting safety and environmental protection. The objective function of the mathematical optimization model of secondary coal blending is established as follows:

Where _Ui is the operating parameter of the ith coal mill, _Ui takes 0 to represent shutdown and takes 1 to represent operation; _Xi is the coal feeding rate of the ith coal mill; _Pi is the unit price of the mixed coal and standard coal in the ith coal bunker; N is the number of coal mills in the unit.

Preferably, the secondary coal blending mathematical optimization model established by the secondary coal blending optimization module based on the coal mill satisfies the following constraints:

(1) The real-time boiler load determined by the output of each coal mill and the coal type information of the corresponding coal bunker must meet the sum of the planned electrical load and the steam and heating load at that time. The load constraints are as follows:

Where, _hi is the calorific value of the mixed coal in the coal bunker corresponding to the i-th coal mill; _Mh is the standard coal consumption under this working condition; _Fd , _Fr , _Fq are the electrical load, steam supply load and heating load of the boiler load respectively;

The sulfur content of the coal entering the furnace every hour must be less than the maximum desulfurization capacity of the power plant at that time. This constraint is:

Where, _si is the sulfur content of the mixed coal in the coal bunker corresponding to the i-th coal mill; _St is the maximum desulfurization capacity of the desulfurization equipment;

In order to avoid slagging in the furnace, the ash content of the coal mixed into the furnace must be limited. This constraint condition is expressed as:

In the formula, a _i is the ash content of the mixed coal in the coal bunker corresponding to the i-th coal mill; A is the total ash content limit of the mixed coal entering the furnace;

The volatile matter of the mixed coal entering the furnace is restricted, and the constraints are as follows:

Where, _vi is the volatile matter of the mixed coal in the coal bunker corresponding to the i-th coal mill; V is the total limit of volatile matter of the mixed coal entering the furnace;

Each coal mill has an upper and lower output limit. In order to make the coal mill operate within the normal output range, the following elastic constraint equation is established:

^In the formula, _Vimax and _Vimin are the maximum and minimum outputs of the ith coal mill, respectively, ^and the elastic range of _Vimin ^is Vimin _is ^the lower ^limit of _Vimin , is ^the upper limit of _Vimin .

Preferably, the coal mill-based secondary coal blending optimization module converts the secondary coal blending mathematical optimization model with elastic constraints into an optimization model with inelastic constraints;

The elastic constraint conditions in the elastic constraint equation (8) are fuzzy, and the membership function of the coal mill coal feeding rate in this interval is defined as follows:

Where A _i ( _xi ) is the fuzzy set of elastic constraints on coal feeding rate of coal mill;

If the elastic constraint of equation (8) is ^When , let z ₀ be the optimal value of the ordinary linear programming corresponding to the optimization model; if the elastic constraint of equation (8) is Vi _{min ≤xi} ≤V _i ^max , let z ₁ be the optimal value of the ordinary linear programming corresponding to it, z ₁ <z ₀ ; let d ₀ =z ₀ -z ₁ , then d ₀ >0, which is the expansion index of the objective function in the fuzzy linear programming; fuzzify the objective function of the fuzzy linear programming, and the standard function membership function is:

In the formula,

The following conclusions can be drawn from the definitions of A _i ( _xi ) and G (t ₀ ):

1) For any λ∈[0,1], we have

In the formula,

2) For any λ∈[0,1], we have

Where G(t ₀ (x)) is the fuzzy set membership function of the objective function;

In order to solve the mathematical optimization model (3) of secondary coal blending and facilitate calculation, let

Where A(x) is the intersection of the fuzzy sets of all coal mill coal feeding rate elastic constraints;

Since symmetric fuzzy discrimination treats the objective function and all constraints equally, all fuzzy constraints should be satisfied as much as possible and the objective function should be optimized as much as possible. It is required that x ^* satisfies _Ai (x ^* ) ≥ λ and G (x ^* ) ≥ λ, and λ reaches the maximum value. Based on conclusions 1 and 2, the mathematical optimization model of secondary coal blending with elastic constraints is converted into the following optimization model with inelastic constraints:

maxλ (14)

Where, other conventional constraints are equations (4), (5), (6), (7);

Assuming the optimal solution of the optimization model (14) with inelastic constraints to be x ^* and λ ^* , the mathematical optimization model of secondary coal blending with elastic constraints can be transformed into a two-stage simplex solution. Equations (3) and (14) are equivalent. The optimal solution of the mathematical optimization model of secondary coal blending with elastic constraints is x ^* , and the optimal value is t=t(x ^* ).

Preferably, the model solving module adopts a two-stage simplex method to solve the primary coal blending mathematical optimization model and the secondary coal blending mathematical optimization model respectively. The specific method is:

Step S1: transforming the primary and secondary coal blending optimization models into standard linear programming models by adding slack variables and residual variables;

Step S2: based on the number of unit bases n (n<m) of the standard linear programming model, an m-order artificial vector u=[u ₁ ,u ₂ ,...,u _mn ] ^T is added to the standard linear programming model as a standard admissible basis;

Step S3: for m-n artificial vectors, the objective function of the standard linear programming model is changed to u ₁ +u ₂ +u ₃ +...+u _mn , and then a one-stage simplex algorithm is used for calculation; Step S4: when the objective function u ₁ +u ₂ +u ₃ +...+u _mn is equal to 0, a basic admissible solution is obtained as a new standard admissible basis, and then the objective function of the original primary coal blending mathematical optimization model or the secondary coal blending mathematical optimization model is solved by the one-stage simplex method to obtain the optimal solution of the primary coal blending optimization model or the secondary coal blending optimization model.

Preferably, an analog data frame is designed between the external coal blending control and regulation system and the original control system to transmit the coal feeding rate optimization value of the coal mill unit, an analog request frame is used to request an analog signal, a mill switching data frame is used to transmit the operation and shutdown signals of the coal mill unit, and a heartbeat switch request frame is used to verify whether the system is online;

The original control system sends a heartbeat switch request frame to the external coal blending control and regulation system. The external coal blending control and regulation system simulates a heartbeat signal through the response of the programmed digital switch signal (0,1). The original control system detects whether the communication of the external system is normal through the heartbeat frame response. When the original control system fails to detect a heartbeat frame response for more than 30s, the external system is automatically cut off and switched to the original control system.

The beneficial effects of adopting the above technical solution are: the coal bunker and coal mill external deep coal blending control system based on the cascade optimization strategy provided by the present invention, (1) the coal blending and combustion process of thermal power enterprises generally relies on the manual experience of experts and combines with the simple calculation mode of Excel table, and rarely adopts optimization technology. The biggest disadvantage of manual experience blending is that it is difficult to ensure the optimization of the mixed coal results. The present invention establishes an automatic one-time coal blending module and a model solving module for the coal bunker based on the optimization model, which greatly improves the coal blending efficiency and significantly reduces the coal blending cost.

(2) In order to address the drawback that coal bunker coal blending does not consider the optimization of coal mill operation combination, the present invention proposes a secondary optimization coal blending module and a model solution module based on the optimization model to select a reasonable coal mill operation combination and the optimal output point, thereby optimizing the simple functional relationship between the coal feeding rate and the load of the original control system, and further saving the coal blending cost.

(3) Aiming at the actual demand for anti-vibration of coal mill, a linear fuzzy programming optimization model with elastic constraints and its model solving module were established. While meeting the safe operation conditions of coal mill, the lower limit of coal feeding rate required by dynamic load was further optimized.

(4) Each module proposed in the present invention is implemented on an external server, which is connected to the original control system by means of communication. The hardware and program of the original control system are not modified. The original manual experience method is only modified into an external computer control and adjustment method based on an optimization algorithm. Heartbeat pulses and disturbance-free switching are used for the external server to ensure system security. Once a communication failure occurs, the original control system can also operate normally.

(5) In order to address the drawbacks of the one-time coal allocation in the coal bunker in front of the furnace without considering the optimization of the coal mill operation combination and the actual needs of variable load, a progressive coal allocation between the coal bunker and the coal mill based on the cascade optimization model was proposed for the first time, providing a new solution for thermal power companies to achieve accurate coal allocation and reduce fuel costs.

(6) Based on the coal blending in the primary coal bunker, the present invention proposes for the first time a secondary optimization coal blending technology based on the optimization model to select a reasonable coal mill operation combination and the optimal output point, thereby optimizing the simple functional relationship between the coal feeding rate and the load of the original control system of the enterprise, and further reducing the fuel cost.

(7) In the secondary optimization of coal blending, taking into account the actual needs of coal mill vibration prevention, a fuzzy linear programming optimization coal blending model with elastic constraints and its solution method were established for the first time. This model can further optimize the ability of the coal mill combination and output point to adapt to variable loads while meeting the safe operation conditions of the coal mill.

The coal bunker and coal mill external coal blending control and adjustment system based on the cascade optimization model proposed in the present invention combines the primary coal blending optimization model, the secondary coal blending optimization model and fuzzy linear programming and other technologies to optimize the coal blending and combustion process of the power plant. In view of the disadvantages of the mixed coal combustion currently commonly used in thermal power plants, which does not consider the operation combination of the coal mill and the optimal output point, the present invention proposes a coal bunker and coal mill external coal blending control and adjustment system based on the cascade optimization model, which optimizes the simple functional relationship between the coal feeding rate and the load of the original control system of the enterprise, and can further reduce fuel costs. In addition, considering the actual needs of the coal mill vibration prevention, a fuzzy linear programming optimization coal blending model with elastic constraints and its solution method are established, which can further optimize the coal mill combination and the ability of the output point to adapt to variable loads while meeting the safe operation conditions of the coal mill.

In summary, the present invention has foreseeable greater economic and social value.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a schematic diagram of a physical architecture of progressive coal blending between a coal bunker and a coal mill based on a cascade optimization model provided by an embodiment of the present invention;

FIG2 is a schematic diagram of the principle and structure of progressive coal blending between a coal bunker and a coal mill based on a cascade optimization model provided by an embodiment of the present invention, wherein (a) is a principle diagram and (b) is a structural diagram;

FIG3 is a fuzzy membership function of the coal feeding rate _xi of the ith coal mill provided by an embodiment of the present invention;

FIG4 is a fuzzy membership function of t ₀ provided by an embodiment of the present invention;

FIG5 is a flow chart of a two-stage simplex solving method provided by an embodiment of the present invention;

6 is a communication flow chart of a coal bunker and a coal mill external deep coal blending control system based on a cascade optimization strategy provided by an embodiment of the present invention;

FIG7 is a diagram showing the implementation architecture of a coal bunker and coal mill external deep coal blending control system based on a cascade optimization strategy provided by an embodiment of the present invention;

FIG8 is a diagram showing an interface of coal blending software based on an optimization model provided by an embodiment of the present invention;

FIG9 is a diagram of a coal blending mill setting interface based on an optimization model provided in an embodiment of the present invention.

DETAILED DESCRIPTION

The specific implementation of the present invention is further described in detail below in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

In this embodiment, the coal bunker and coal mill external deep coal blending control system based on the cascade optimization strategy, as shown in Figures 1 and 2, includes a server, a communication module, and a primary optimization coal blending module based on the coal bunker, a secondary optimization coal blending module based on the coal mill, a model solving module, a relational database interface module, and a real-time database interface module formed by software programming and set on the server;

The relational database interface module is established based on ADO.NET, and reads the corresponding coal quality parameters from the SQL Server relational database storing the coal quality parameter information (such as calorific value, sulfur, moisture, volatile matter, ash content) of each single type of coal in the coal yard, so as to provide the basic coal storage data of the coal yard for the first optimization of coal blending;

The real-time database interface module is established based on OPAPI.NET, and reads the corresponding real-time data of operating parameters from the OPENPlant real-time database storing real-time data of unit operating parameters (such as load, coal mill coal feeding rate, coal level meter, etc.), providing the real-time data basis of the unit for secondary optimization of coal blending;

The coal bunker-based one-time optimization coal blending module aims to pursue the optimal coal blending cost of the coal bunker under the premise of ensuring the safety and environmental protection of the generator set and driving the maximum planned load of the boiler (power generation load, heating load and steam supply load). In the coal bunker mixed coal blending stage, the coal type and proportion of each coal bunker are used as optimization variables to establish a one-time coal blending mathematical optimization model; at the same time, the one-time optimization coal blending module transmits the mixed coal blending coal type and proportion of each coal bunker to the upper computer of the original reclaimer control system through the communication module (network communication based on Ethernet), and coordinates the coordinated coal loading of the two reclaimers;

The secondary coal blending module based on the coal mill, based on the primary optimization coal blending, ensures the safety and environmental protection of the generator set and drives the dynamic planned load (not the maximum planned load) of the boiler under the constraint of pursuing the optimal coal mill operation cost. In the mixed coal operation stage of the coal mill, a secondary coal blending mathematical optimization model is established with the operation combination and optimal output (coal feeding rate) of each coal mill as the optimization variables; at the same time, the secondary optimization coal blending module transmits the mill operation plan of the optimal combination and optimal output point of the coal mill through the communication module (serial port-based communication) in the RTU mode based on the 485 bus and the Modbus protocol, so as to realize the compensation adjustment of the coal feeding rate of the original coal mill control system.

The model solving module, based on the established mathematical optimization model of primary coal blending and the mathematical optimization model of secondary coal blending, realizes modeling and solving of the mathematical optimization model of primary coal blending and the mathematical optimization model of secondary coal blending through software programming, solves the mixed coal blending scheme of the coal bunker, the optimal combination of coal mills and the mill operation scheme of the optimal output point, and then realizes the control and adjustment of the coal bunker and the coal mill;

The coal bunker-based primary coal blending optimization module takes the mixed coal combustion meeting the calorific value, sulfur, volatile matter and ash content indicators as constraint conditions, and takes the economic index of the minimum mixed coal as the objective function to establish a primary coal blending mathematical optimization model, as shown in the following formula:

In the formula, minf(x) is the minimum cost of mixed coal, _cj is the cost of the j-th single coal, that is, the unit price of standard coal, _xj is the blending ratio of the j-th single coal, and n is the total number of coal types in the mixed coal;

According to the actual situation of the enterprise's equipment and coal types, some coal quality parameter characteristic indicators of mixed coal are selected as constraints; assuming that the enterprise has restrictions on the actual calorific value, actual sulfur content, actual ash content and actual volatile matter of the mixed coal entering the furnace, the constraints of mixed coal are:

In the formula, _hj , _sj , _aj , _vj represent the calorific value, sulfur content, ash content and volatile matter of each type of coal respectively; H, S, A, V represent the upper or lower limits of the calorific value, sulfur content, ash content and volatile matter of the mixed coal scheme respectively; according to the actual situation of the enterprise, the constraints can also include slagging characteristics, burnout characteristics, grindability and other coal quality characteristics indicators as well as other special constraints of the enterprise's own characteristics and needs (such as the priority of special coal types, ash melting point, virtual coal types). After establishing the pre-furnace coal blending optimization model, the optimization algorithm is used to solve the coal blending scheme for the coal bunker at one time, and based on this scheme, a bunkering plan is formulated, and then coal is added to each coal bunker to complete the coal blending task at one time.

In this embodiment, for the optimization of coal blending at a high load, 7 types of coal are piled in the experimental coal yard, and each coal bunker does not exceed 2 types of coal. The high load of power generation is 310MW, and the heating load and steam supply load are 10MW respectively. Based on the maximum desulfurization capacity of the unit, it can be calculated that the upper limit of the total sulfur of the unit's optimization coal blending model is 0.87, and the upper limits of volatile matter and ash are set to 35 and 25. At this time, the boiler load (330MW) is a high-load operation mode, and the coal mills are running A, B and D. The maximum coal feeding rate of each coal mill is shown in Table 1. The optimal high-load coal blending scheme for the ABD coal bunker in Table 2 can be obtained by using the optimization model of formulas (1) and (2) and its constraints and solving it through the two-stage simplex algorithm.

Based on the above high-load optimization coal allocation, the C mill is configured for medium-high load of 210MW, heating load and steam load are also 10MW respectively, and the corresponding constraint parameters are consistent with the high load. At this time, the boiler load is 230MW, which is the medium-high load operation mode, and the coal mills are running A, B and C. The same optimization model of formulas (1) and (2) and the two-stage simplex algorithm are used to obtain the optimal medium-high load coal allocation plan for ABC coal bunker in Table 2.

Table 1 Coal mill and parameter configuration

Coal mill serial number A Grinding B Grinding C Grinding D Grinding Coal mill function Frequent transfer Frequent transfer Medium and high spare High load standby Maximum output (T/h) 52 54 53 55 Minimum output range (T/h) 26-30 26-30 26-30 26-30

Table 2 Actual primary coal blending scheme for thermal power plants

The secondary coal blending optimization module based on coal mill takes the real-time fuel cost as the target, introduces the operating parameter variables of each coal mill in combination with the actual situation, and makes the real-time fuel cost of power production the lowest under the premise of meeting safety and environmental protection. The objective function of the mathematical optimization model of secondary coal blending is established as follows:

Where _Ui is the operating parameter of the ith coal mill, where 0 represents shutdown and 1 represents operation; _Xi is the coal feeding rate of the ith coal mill; _Pi is the unit price of the mixed coal and standard coal in the ith coal bunker; N is the number of coal mills in the unit.

The secondary coal blending mathematical optimization model established based on the secondary coal blending optimization module of the coal mill satisfies the following constraints:

(1) The real-time boiler load determined by the output of each coal mill and the coal type information of the corresponding coal bunker must meet the sum of the planned electrical load and the steam and heating load at that time. The load constraints are as follows:

Where, _hi is the calorific value of the mixed coal in the coal bunker corresponding to the i-th coal mill; _Mh is the standard coal consumption under this working condition; _Fd , _Fr , _Fq are the electrical load, steam supply load and heating load of the boiler load respectively;

The environmental protection requirements cannot be ignored in power production. Strict requirements must be imposed on the sulfur content. The sulfur content of the coal entering the furnace every hour must be less than the maximum desulfurization capacity of the power plant at that time. The constraint condition is:

Where, _si is the sulfur content of the mixed coal in the coal bunker corresponding to the i-th coal mill; _St is the maximum desulfurization capacity of the desulfurization equipment;

In order to avoid slagging in the furnace, the ash content of the coal mixed into the furnace must be limited. This constraint condition is expressed as:

In the formula, a _i is the ash content of the mixed coal in the coal bunker corresponding to the i-th coal mill; A is the total ash content limit of the mixed coal entering the furnace;

In order to ensure the safety of combustion in the furnace and avoid explosions in the furnace, the volatile matter of the mixed coal entering the furnace is restricted. The constraints are as follows:

Where, _vi is the volatile matter of the mixed coal in the coal bunker corresponding to the i-th coal mill; V is the total limit of volatile matter of the mixed coal entering the furnace;

Each coal mill has an upper and lower output limit. In order to make the coal mill operate within the normal output range, the following elastic constraint equation is established:

^In the formula, _Vimax and _Vimin are the maximum and minimum outputs of the ith coal mill, respectively, ^and the elastic range of _Vimin ^is Vimin _is ^the lower ^limit of _Vimin , is the upper limit of _Vimin ^;

Considering that too low coal feeding rate of coal mill will cause mill vibration, in order to avoid this hazard, equation (8) is a fuzzy constraint condition, indicating that the constraint has a certain elasticity. The elastic range on the left side of the constraint is If the coal feeding rate of the coal mill is greater than or equal to When the engineer is "satisfied", When it is between , the engineer's satisfaction gradually decreases, _and when it is less than Vi ^min , the engineer is dissatisfied.

The coal mill-based secondary coal blending optimization module converts the secondary coal blending mathematical optimization model with elastic constraints into an optimization model with inelastic constraints;

Since equation (8) contains elastic constraints, the traditional two-stage simplex method cannot be directly solved. Therefore, the elastic constraints in equation (8) are fuzzified, and the membership function of the coal mill coal feeding rate in this interval is defined as follows:

Where A _i ( _xi ) is the fuzzy set of elastic constraints on coal feeding rate of coal mill, and the elastic constraint membership function of _xi is shown in Figure 3.

If the elastic constraint of equation (8) is When , let z ₀ be the optimal value of the ordinary linear programming corresponding to the optimization model; if the elastic constraint of equation (8) is Vi _{min ≤xi} ≤V _i ^max , let z ₁ be the optimal value of the ordinary linear programming corresponding to it ^. Generally speaking, z ₁ <z ₀ should be satisfied, otherwise the expansion index of the fuzzy constraint condition in the fuzzy linear programming will not work. Let d ₀ =z ₀ -z ₁ , then d ₀ >0, which is the expansion index of the objective function in the fuzzy linear programming; fuzzify the objective function of the fuzzy linear programming, and the membership function of the label function is:

In the formula, The graph is shown in Figure 4.

The following conclusions can be drawn from the definitions of A _i ( _xi ) and G (t ₀ ):

1) For any λ∈[0,1], we have

In the formula,

2) For any λ∈[0,1], we have

Where G(t ₀ (x)) is the fuzzy set membership function of the objective function;

In order to solve the mathematical optimization model (3) of secondary coal blending and facilitate calculation, let

Where A(x) is the intersection of the fuzzy sets of elastic constraints on the coal feeding rates of all coal mills. In this example, m=4 refers to mills A, B, C, and D.

Since symmetric fuzzy discrimination treats the objective function and all constraints equally, all fuzzy constraints should be satisfied as much as possible and the objective function should be optimized as much as possible. It is required that x ^* satisfies _Ai (x ^* ) ≥ λ and G (x ^* ) ≥ λ, and λ reaches the maximum value. Based on conclusions 1 and 2, the mathematical optimization model of secondary coal blending with elastic constraints is converted into the following optimization model with inelastic constraints:

maxλ (14)

Where, other conventional constraints are equations (4), (5), (6), (7);

Assuming the optimal solution of the optimization model (14) with inelastic constraints to be x ^* and λ ^* , the mathematical optimization model of secondary coal blending with elastic constraints can be transformed into a two-stage simplex solution. (3) and (14) are equivalent, and the optimal solution is x ^* , and the optimal value is t=t(x ^* ).

The above coal blending scheme based on high load and medium-high load is carried out under the condition of maximum coal feeding rate of the coal mill. As the load changes, the coal feeding rate changes with the load. On the basis of primary coal blending, there are few studies on secondary coal blending to find the optimal mill combination and the optimal matching coal feeding rate. This embodiment gives the optimal rate combination based on the coal mill optimization model (3) to (14) at a low load of 130MW, see the low load output in Table 3. From the results of the secondary optimization coal blending experiment, it can be concluded that the coal feeding rates of mills A, B and C are 28.00T/h, 29.08T/h and 28.54T/h respectively, all within the fuzzy interval of 26-30, making the coal blending and combustion scheme more optimized and economical, and also within the allowable range to avoid vibration.

Table 3 Secondary optimization scheme of coal mill combination under fuzzy decision making

Coal mill serial number A Grinding B Grinding C Grinding D Grinding Objective Function High load output (T/h) 52.00 54.00 —— 55.00 175458 Medium and high load output (T/h) 52.00 54.00 53.00 —— 166139 Low load output (T/h) 28.00 29.08 28.54 —— 89464

Due to the needs of actual engineering problems, there are inequality constraints and equality constraints in the first coal blending optimization model and the second coal blending optimization model established above. The first and second coal blending optimization models can be transformed into standard linear programming models by adding slack variables and residual variables. Since the constraint matrix in the converted standard linear programming model does not contain an m-order (m is the number of constraint equations) unit matrix, it means that there is no initial admissible basis at this moment. In order to make the constraint matrix have an m-order unit matrix, that is, to have an m-order standard admissible basis, based on the number of standard model unit bases n (n<m), a new artificial vector u＝[u ₁ ,u ₂ ,...,u _mn ] ^T is introduced, and then the following two-stage simplex method is used to solve.

Therefore, the model solving module adopts the two-stage simplex method to solve the mathematical optimization model of primary coal blending and the mathematical optimization model of secondary coal blending respectively. The specific method is:

Step S1: transforming the primary and secondary coal blending optimization models into standard linear programming models by adding slack variables and residual variables;

Step S2: based on the number of unit bases n (n<m) of the standard linear programming model, an m-order artificial vector u=[u ₁ ,u ₂ ,...,u _mn ] ^T is added to the standard linear programming model as a standard admissible basis;

Step S3: for m-n artificial vectors, the objective function of the standard linear programming model is changed to u ₁ +u ₂ +u ₃ +...+ _umn , and then the one-stage simplex algorithm is used for calculation; this process is the first stage of the two-stage simplex method;

Step S4: When the objective function _u1 + _u2 + _u3 +...+ _umn is equal to 0 (if it is not equal to 0, there is no solution), a basic admissible solution is obtained as a new standard admissible basis, and then the objective function of the original primary coal blending mathematical optimization model or the secondary coal blending mathematical optimization model is solved by the one-stage simplex method. This process is the second stage of the two-stage simplex method, and the optimal solution of the primary coal blending optimization model or the secondary coal blending optimization model is obtained. The above solution process is shown in Figure 5.

The external coal blending control and regulation system and the original control system are designed with analog data frames for transmitting the coal feeding rate optimization value of the coal mill unit, analog request frames for requesting analog signals, mill switching data frames for transmitting the operation and shutdown signals of the coal mill unit, and heartbeat switch request frames for verifying whether the system is online;

In order to ensure the safety of system operation and not affect production requirements, the original control system sends a heartbeat switch quantity request frame to the external coal blending control and regulation system. The external coal blending control and regulation system simulates the heartbeat signal through the response of the programmed digital switch signal (0,1), and the original control system detects whether the communication of the external system is normal through the heartbeat frame response; when the original control system fails to detect the heartbeat frame response for more than 30s, it automatically cuts off the external system and switches to the original control system, ensuring the safety and normal operation of the original control system of the power plant;

The external coal blending control and regulation system calculates the coal type ratio of the coal bunker mixed coal, the optimal combination and output point of the coal mill based on the primary coal blending module, secondary coal blending module and solution module of the cascade optimization model, and transmits them to the original control system through the communication module to realize the automatic adjustment of the bunker addition, coal mill combination and coal feeding rate. Considering the communication task processing capability, the coal mill coal feeding rate adjustment control cycle is set to 15s to ensure that the external coal blending control and regulation system has sufficient time to complete the online calculation and process adjustment. The above communication process is shown in Figure 6.

In this embodiment, two stackers, two loading conveyors, four coal bunkers and four coal mills are coordinated, optimized and controlled through the reclaimer control system and the coal mill control system. Based on the control network and Ethernet network and supported by the relational database and real-time database, the external coal blending control and adjustment system is physically independent of the original control system under safe conditions without affecting the production of the power plant. The external coal blending control and adjustment system is connected to the original control system by communication, and the hardware and program of the original control system are not changed. It is just modified from the original manual experience method to an automatic implementation method of an external computer based on an algorithm module. The external coal blending control and adjustment system uses heartbeat pulses and disturbance-free switching to ensure system safety. Once a communication failure occurs, the original control system can operate normally and safely.

Implementation of one-time coal blending: 1) Read the maximum load, maximum medium and high load and constraint boundary conditions of the boiler through the relational database interface module; 2) Read the coal quality parameters of each single coal type in the relational database through the interface module; 3) Based on the working range of the two reclaimers, the coal type combination of each coal yard is circulated through the coal yard matching module; 4) Based on the one-time optimization model module and the solution module, the coal type and proportion of the mixed coal in each coal bunker are calculated; 5) The external system coordinates and adjusts the set value of the reclaimer control system through communication. The reclaimer control system uses PID algorithm to control the coordinated work of the two reclaimers, loads coal into the coal bunker in proportion, and completes a coal blending task.

The realization of secondary coal blending: 1) read the coal bunker coal blending plan of the relational database through the real-time database interface module; 2) read the real-time load of the real-time database in different time periods through the interface module; 3) calculate the optimal combination and output point of the coal mill through the secondary optimization model and solution module; 4) switch the operation status of the mill according to the optimal mill combination (implemented after confirmation by the operating personnel); 5) the external system is connected to the coal mill control system by communication, and the coal feeding rate (output) of the coal mill is adjusted by the offset to complete the secondary coal blending task. The system architecture and interface are shown in Figures 7-9.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some or all of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope defined by the claims of the present invention.

Claims (7)

1. A coal bunker and coal mill external deep coal blending control system based on cascade optimization strategy, characterized by: comprising a server, a communication module, and a primary coal blending module based on the coal bunker, a secondary coal blending module based on the coal mill, a model solving module, a relational database interface module and a real-time database interface module formed by software programming and set on the server; The relational database interface module reads the corresponding coal quality parameters from the SQL Server relational database storing the coal quality parameter information of each single coal type in the coal yard, and provides the basic coal storage data in the coal yard for the primary optimization of coal blending; The real-time database interface module reads the corresponding real-time data of operating parameters from the OPENPlant real-time database storing the real-time data of the unit operating parameters, and provides the real-time data basis of the unit for the secondary optimization of coal blending; The coal bunker-based one-time optimization coal blending module establishes a one-time coal blending mathematical optimization model with the coal types and proportions of each coal bunker as optimization variables during the mixed coal blending stage of the coal bunker; at the same time, the one-time optimization coal blending module transmits the mixed coal types and proportions of each coal bunker to the upper computer of the original reclaimer control system through the communication module, and coordinates the coordinated coal loading of the two reclaimers; The secondary coal blending optimization module based on the coal mill, based on the primary coal blending optimization, establishes a secondary coal blending mathematical optimization model with the operation combination and optimal output of each coal mill as optimization variables during the coal mixing operation stage of the coal mill; at the same time, the secondary coal blending optimization module transmits the coal mill operation plan of the optimal combination and optimal output point of the coal mill through the communication module, so as to realize the compensation adjustment of the coal feeding rate of the original coal mill control system; The model solving module, based on the established primary coal blending mathematical optimization model and secondary coal blending mathematical optimization model, realizes the modeling and solving of the primary coal blending mathematical optimization model and the secondary coal blending mathematical optimization model, solves the mixed coal blending scheme of the coal bunker, the optimal combination of coal mills and the mill operation scheme of the optimal output point, and then realizes the control and adjustment of the coal bunker and the coal mill. 2. According to the coal bunker and coal mill external deep coal blending control system based on cascade optimization strategy in claim 1, it is characterized in that: the coal bunker-based primary coal blending optimization module takes the mixed coal combustion to meet the calorific value, sulfur, volatile matter and ash content indicators as constraint conditions, and takes the economic index of the minimum mixed coal as the objective function, and establishes a primary coal blending mathematical optimization model, as shown in the following formula: In the formula, min f(x) is the minimum cost of mixed coal, c _j is the cost of the j-th single coal, that is, the unit price of standard coal, x _j is the blending ratio of the j-th single coal, and n is the total number of coal types in the mixed coal; According to the actual situation of the enterprise's equipment and coal types, some coal quality parameter characteristic indicators of mixed coal are selected as constraints; assuming that the enterprise has restrictions on the actual calorific value, actual sulfur content, actual ash content and actual volatile matter of the mixed coal entering the furnace, the constraints of mixed coal are: In the formula, _hj , _sj , _aj , and _vj represent the calorific value, sulfur content, ash content, and volatile matter of each type of coal respectively; H, S, A, and V represent the upper or lower limits of the calorific value, sulfur content, ash content, and volatile matter of the mixed coal scheme respectively; according to the actual situation of the enterprise, coal quality characteristic indicators and other special constraints of the enterprise's own characteristics and needs can also be added to the constraints. 3. According to the coal bunker and coal mill external deep coal blending control system based on cascade optimization strategy in claim 2, it is characterized in that: the secondary coal blending optimization module based on coal mill takes real-time fuel cost as the target, introduces the operating parameter variables of each coal mill in combination with the actual situation, and makes the real-time fuel cost of power production the lowest under the premise of meeting safety and environmental protection, and the objective function of the mathematical optimization model of secondary coal blending is established as follows: Where _Ui is the operating parameter of the ith coal mill, _Ui takes 0 to represent shutdown and takes 1 to represent operation; _Xi is the coal feeding rate of the ith coal mill; _Pi is the unit price of the mixed coal and standard coal in the ith coal bunker; N is the number of coal mills in the unit. 4. The coal bunker and coal mill external deep coal blending control system based on cascade optimization strategy according to claim 3 is characterized in that the secondary coal blending mathematical optimization model established by the secondary coal blending optimization module based on the coal mill satisfies the following constraints: (1) The real-time boiler load determined by the output of each coal mill and the coal type information of the corresponding coal bunker must meet the sum of the planned electrical load and the steam and heating load at that time. The load constraints are as follows: Where, _hi is the calorific value of the mixed coal in the coal bunker corresponding to the i-th coal mill; _Mh is the standard coal consumption under this working condition; _Fd , _Fr , _Fq are the electrical load, steam supply load and heating load of the boiler load respectively; The sulfur content of the coal entering the furnace every hour must be less than the maximum desulfurization capacity of the power plant at that time. This constraint is: Where, _si is the sulfur content of the mixed coal in the coal bunker corresponding to the i-th coal mill; _St is the maximum desulfurization capacity of the desulfurization equipment; In order to avoid slagging in the furnace, the ash content of the coal mixed into the furnace must be limited. This constraint condition is expressed as: In the formula, a _i is the ash content of the mixed coal in the coal bunker corresponding to the i-th coal mill; A is the total ash content limit of the mixed coal entering the furnace; The volatile matter of the mixed coal entering the furnace is restricted, and the constraints are as follows: Where, _vi is the volatile matter of the mixed coal in the coal bunker corresponding to the i-th coal mill; V is the total limit of volatile matter of the mixed coal entering the furnace; Each coal mill has an upper and lower output limit. In order to make the coal mill operate within the normal output range, the following elastic constraint equation is established: ^In the formula, _Vimax and _Vimin are the maximum and minimum outputs of the ith coal mill, respectively, ^and the elastic range of _Vimin ^is Vimin _is ^the lower ^limit of _Vimin , is ^the upper limit of _Vimin . 5. The coal bunker and coal mill external deep coal blending control system based on cascade optimization strategy according to claim 4 is characterized in that: the secondary coal blending optimization module based on the coal mill converts the secondary coal blending mathematical optimization model with elastic constraints into an optimization model with inelastic constraints; The elastic constraint conditions in the elastic constraint equation (8) are fuzzy, and the membership function of the coal mill coal feeding rate in this interval is defined as follows: Where A _i ( _xi ) is the fuzzy set of elastic constraints on coal feeding rate of coal mill; If the elastic constraint of equation (8) is ^When , let z ₀ be the optimal value of the ordinary linear programming corresponding to the optimization model; if the elastic constraint of equation (8) is Vi _{min ≤xi} ≤V _i ^max , let z ₁ be the optimal value of the ordinary linear programming corresponding to it, z ₁ <z ₀ ; let d ₀ =z ₀ -z ₁ , then d ₀ >0, which is the expansion index of the objective function in the fuzzy linear programming; fuzzify the objective function of the fuzzy linear programming, and the standard function membership function is: In the formula, The following conclusions can be drawn from the definitions of A _i ( _xi ) and G (t ₀ ): 1) For any λ∈[0,1], we have In the formula, 2) For any λ∈[0,1], we have Where G(t ₀ (x)) is the fuzzy set membership function of the objective function; In order to solve the mathematical optimization model (3) of secondary coal blending and facilitate calculation, let Where A(x) is the intersection of the fuzzy sets of all coal mill coal feeding rate elastic constraints; Since symmetric fuzzy discrimination treats the objective function and all constraints equally, all fuzzy constraints should be satisfied as much as possible and the objective function should be optimized as much as possible. It is required that x ^* satisfies _Ai (x ^* ) ≥ λ and G (x ^* ) ≥ λ, and λ reaches the maximum value. Based on conclusions 1 and 2, the mathematical optimization model of secondary coal blending with elastic constraints is converted into the following optimization model with inelastic constraints: maxλ (14) Where, other conventional constraints are equations (4), (5), (6), (7); Assuming the optimal solution of the optimization model (14) with inelastic constraints to be x ^* and λ ^* , the mathematical optimization model of secondary coal blending with elastic constraints can be transformed into a two-stage simplex solution. Equations (3) and (14) are equivalent. The optimal solution of the mathematical optimization model of secondary coal blending with elastic constraints is x ^* , and the optimal value is t=t(x ^* ). 6. The coal bunker and coal mill external deep coal blending control system based on cascade optimization strategy according to claim 5 is characterized in that: the model solving module adopts a two-stage simplex method to solve the primary coal blending mathematical optimization model and the secondary coal blending mathematical optimization model respectively. The specific method is: Step S1: transforming the primary and secondary coal blending optimization models into standard linear programming models by adding slack variables and residual variables; Step S2: based on the number of unit bases n (n<m) of the standard linear programming model, an m-order artificial vector u=[u ₁ ,u ₂ ,...,u _mn ] ^T is added to the standard linear programming model as a standard admissible basis; Step S3: for m-n artificial vectors, the objective function of the standard linear programming model is changed to u ₁ +u ₂ +u ₃ +...+u _mn , and then a one-stage simplex algorithm is used for calculation; Step S4: when the objective function u ₁ +u ₂ +u ₃ +...+u _mn is equal to 0, a basic admissible solution is obtained as a new standard admissible basis, and then the objective function of the original primary coal blending mathematical optimization model or the secondary coal blending mathematical optimization model is solved by the one-stage simplex method to obtain the optimal solution of the primary coal blending optimization model or the secondary coal blending optimization model. 7. The coal bunker and coal mill external deep coal blending control system based on cascade optimization strategy according to claim 6 is characterized in that: an analog data frame is designed between the external coal blending control and adjustment system and the original control system to transmit the coal feeding rate optimization value of the coal mill unit, an analog request frame is used to request an analog signal, a mill switching data frame is used to transmit the operation and shutdown signal of the coal mill unit, and a heartbeat switch request frame is used to verify whether the system is online; The original control system sends a heartbeat switch request frame to the external coal blending control and regulation system. The external coal blending control and regulation system simulates a heartbeat signal through the response of the programmed digital switch signal (0,1). The original control system detects whether the communication of the external system is normal through the heartbeat frame response. When the original control system fails to detect a heartbeat frame response for more than 30s, the external system is automatically cut off and switched to the original control system.