CN110471285A - Based on the event driven zinc abstraction roasting process fuzzy control method of trend - Google Patents

Abstract

The present invention is a kind of based on the event driven zinc abstraction roasting process fuzzy control method of trend, pass through set temperature setting value and control period N, according to the real-time sampling temperature value of sensor, calculate temperature deviation, Extracting temperature trend, and when changing in operating condition or reaching the preset control period, fuzzy control is carried out according to temperature trend and its temperature deviation in time, improves the problem of operating condition assessment difficulty and fuzzy Control performance caused by being constrained the dynamic characteristic and site environment of roasting process are declined.

Description

Fuzzy control method for zinc smelting and roasting process based on trend event drive

technical field

The invention relates to the technical field of fuzzy control, in particular to a trend event-driven fuzzy control method for zinc smelting and roasting process.

Background technique

The roasting process is the first process in the zinc smelting process. In this process, the mixed zinc concentrate is sent to the roaster for full combustion, and zinc calcine, sulfur dioxide, smoke and other products are produced. Zinc calcine is the main raw material in the zinc hydrometallurgy process, and its product quality is crucial to the production of downstream processes. The main purpose of the roasting process is to ensure the product quality of the zinc calcine, that is, to increase the soluble zinc rate of the zinc calcine and reduce the content of insoluble impurities. Since the quality of zinc calcine mainly depends on the composition of the mixed zinc concentrate and the temperature of the roasting process, the most important problem in the roasting process is to ensure the stability of the roasting temperature under different working conditions. Due to the complex dynamic characteristics of the roasting process and the on-site environment, various working conditions often change during the roasting process. It is difficult for the traditional PID controller to achieve stable temperature control in the industrial site. However, the performance of manual control based on experience is greatly dependent on the operator's subjective factors, resulting in unstable performance of manual control, often untimely and inappropriate control.

Based on the characteristics of the roasting process, fuzzy control is more suitable for the temperature control in the roasting process. Most of the fuzzy control of the temperature in the prior art uses the temperature deviation and the change of the temperature deviation as input variables, indicating the current state and the future temperature in a certain period of time respectively. Potential state, according to these two states, the working conditions are evaluated, and then the adjustment value of the output is controlled, that is, the adjustment value of the feed amount, so as to stabilize the temperature within the required range. However, due to the dynamic characteristics of the roasting process and the constraints of the on-site environment, the general control period in the industrial site is selected to be more than ten minutes, which is much longer than the sampling period once per minute. Due to changes in working conditions and various disturbances in the roasting process, inaccurate or erroneous evaluation results may be obtained, resulting in a decline in the performance of the fuzzy controller; if the control cycle is shortened, due to the large It takes a while to reflect the temperature change after the feed amount is adjusted, so frequent adjustments to the feed amount will lead to instability of the roasting system, and the performance of the fuzzy controller will also decline. Therefore, there is a need for a fuzzy control method that can solve the problems of difficult evaluation of working conditions and decreased control performance of fuzzy controllers due to the dynamic characteristics of the roasting process and site environmental constraints.

Contents of the invention

(1) Technical problems to be solved

Based on the above problems, the present invention provides a fuzzy control method for the zinc smelting and roasting process based on trend events, which responds to the dynamic characteristics of the roasting process and the on-site environment, and uses the trend extraction method to carry out effective working condition evaluation. When the working condition changes or reaches During the preset control period, the fuzzy control is carried out in time according to the temperature deviation and temperature trend, so as to improve the performance of the fuzzy controller.

(2) Technical solutions

Based on the above technical problems, the present invention provides a fuzzy control method for zinc smelting and roasting process driven by trend events, the control method includes the following steps:

S1. Set the temperature setting value and control cycle N;

S2. According to the set value and the real-time sampling temperature value of the sensor, calculate the temperature deviation, extract the temperature trend, and screen out the temperature trend and temperature deviation when the working condition changes or reaches the preset control period based on the event-driven strategy of the temperature trend. Timely input to the fuzzy control unit;

S3. Based on the rule setting based on expert experience, the input temperature deviation and temperature trend are fuzzified. After fuzzy reasoning, the adjustment value of the feed amount is output after defuzzification, and the feed amount of the roasting process is adjusted by the actuator.

Further, the event-driven strategy based on temperature trend described in step S2 includes the following steps:

S2.1. Input the given control cycle N, the temperature data within the time window (t ₁ , t _i );

S2.2. Build a trend model;

S2.3. Determine whether the temperature trend has changed. If not, proceed to step S2.4. If it has changed, proceed to step S2.6;

S2.4, judge whether i+1=N, if the result is yes, enter step S2.6;

S2.5. If the result of S2.4 is no, return to step S2.2;

S2.6. Calculate the first-order derivative of the trend model at the current moment;

S2.7. Send the first derivative and temperature deviation at the current moment to the fuzzy controller;

S2.8. Determine whether to end the control, if yes, pause, if not, start a new time window from time t _i+1 , and go to step S2.2.

Further, judging whether the temperature trend described in step S2.3 has changed includes the following steps:

S2.3.1. Calculate the prediction error e _i+1 and the threshold value th _1,i of the first category, and judge whether |e _i+1 |≤th _1,i , if yes, proceed to step S2.3.4;

S2.3.2. If the result of S2.3.1 is no, then judge whether the current temperature y _i+1 is an abnormal value, that is, for the time window (t _i+1 , ) for all moments t _j , i+1≤j≤i+l _th , calculate the corresponding prediction error e _j and the first-class threshold th _1,j-1 , and judge whether all |e _j |≥ th _{1, j-1} , if not, then y _i+1 is an outlier value, enter step S2.3.4;

S2.3.3. If the result of S2.3.2 is yes, the current temperature y _i+1 is not an abnormal value, and the temperature trend has changed;

S2.3.4. Calculate the cumulative error cusum(t _i+1 ) and the second-type threshold th _2,i . Determine whether |cusum(t _i+1 )|≤th _2,i , if yes, the temperature trend has not changed; if not, the temperature trend has changed.

Further, the length of the time window (t ₁ , t _i ) described in step S2.1 is i, l _th ≤ i ≤ N, the control period N is the maximum length of the time window, and l _th is the minimum time window length.

Further, the trend model described in step S2.2 is:

Among them, Y _i =[y ₁ ,y ₂ ...y _i ] ^T , are the parameters of the model, is an estimate of the noise bias, is the output of the model, that is, the temperature prediction value, y _k is the temperature detection value, t _k is the time value, 1≤k≤i≤N, 1≤m≤3.

Further, the prediction error e _i+1 , the first-type threshold th _1,i , the cumulative error cusum(t _i+1 ) and the second-type threshold th _2,i are respectively:

cusum(t _i+1 )＝cusum(t _i )+e _i+1

a _i ＝((T _i ^T T _i ) ^-1 (T _i ) ^T ) ^T [1 (t _i+1 -t ₁ ) (t _i+1 -t ₁ ) ² ] ^T

b _i+1 ＝[(a _i +b _i ) ^T 1] ^T

in, Represents the t distribution, α is the confidence level of the t distribution, Further, the first-order derivative of the current trend model described in step S2.6 is:

Further, the blurring described in step S3 is described as:

The fuzzy description of temperature deviation is: very high VH, slightly high LH, suitable for Z, slightly low LL and very low VL; for VH, use S-type membership function to describe, for VL, use Z-type membership function, and others Then use the bell-shaped membership function to describe;

The fuzzy description of the temperature trend is: very large HP, slightly large LP, stable S, slightly small LN and very small HN; for HP, use S-type membership function to describe, for HN, use Z-type membership function, and others Then use the bell-shaped membership function to describe;

The fuzzy description of the adjustment value of the feed amount is: positive large PB, positive medium PM, positive small PS, zero O, negative small NS, negative medium NM, and negative large NB; PB is described by S-type membership function, and NB is described by Z-shaped membership function, and others are described by bell-shaped membership function.

Further, the Z-type membership function Z ^p (x), the S-type membership function S ^p (x) and the bell-type membership function The rule settings based on expert experience are:

Among them, a ^p , b ^p , c ^p , d ^p , and ν are the parameter values of the membership function, p=1,2,3; q=-2,...,2, p=1,2,3 means that the membership function belongs to the temperature deviation, temperature trend and feed Quantity; q represents different cases of the bell-shaped function.

Further, the deblurring method described in step S3 is the center of gravity method, and its expression is:

Among them, u ^* is the adjustment value of the feed amount, that is, the output value of the fuzzy control, u _f is the adjustment value of the feed amount corresponding to the membership function, n is the number of the membership function of the adjustment value of the feed amount, n=7.

(3) Beneficial effects

The technical scheme of the present invention has the following advantages:

(1) The present invention obtains the temperature trend in real time, utilizes the trend-driven strategy based on the temperature trend, and performs fuzzy control in time according to the temperature trend and temperature deviation when the working condition changes or reaches the preset control cycle, and then controls the feed amount Make adjustments to stabilize the system temperature, enrich the sampling data, and filter the input data of fuzzy control to make the whole control process more timely, more accurate and the system more stable;

(2) The trend-driven strategy based on the temperature trend of the present invention considers the temperature trend change, the temperature trend does not change and the temperature gradually shifts two situations, the working condition evaluation is more accurate, and the fuzzy control is adjusted to deal with the change of the working condition. The influence of the system makes the control performance of fuzzy control better;

(3) The input quantity of fuzzy control of the present invention adopts temperature deviation and temperature trend, rather than the increment of temperature deviation and deviation, accords with the dynamic characteristic of roasting process, has smaller overshoot, and adjustment time is faster, and can be used at work When the working condition changes, the adjustment time is faster, the temperature control is more stable, and the influence of the working condition change is smaller;

(4) The present invention judges the abnormal value, which reduces the possibility of misjudgment;

(5) The present invention uses fuzzy control, which makes up for the deficiency of the mathematical model to a certain extent, and the defuzzification of the method adopts the center of gravity method, which is more intuitive and reasonable.

Description of drawings

The features and advantages of the present invention will be more clearly understood by referring to the accompanying drawings, which are schematic and should not be construed as limiting the invention in any way. In the accompanying drawings:

Fig. 1 is the structural block diagram of the fuzzy control method based on the zinc smelting and roasting of the zinc smelting roasting driven by the trend event in the embodiment of the present invention;

FIG. 2 is a schematic flowchart of an event-driven strategy based on temperature trends in an embodiment of the present invention;

Fig. 3 is the membership function table of the adjustment value of temperature deviation and temperature trend and feed amount of the present invention;

Fig. 4 is the fuzzy inference rule between the adjustment value of temperature deviation and temperature trend and feed amount of the present invention;

Fig. 5 is a control performance comparison chart of Embodiment 1 of the present invention;

FIG. 6 is a comparison table of control performance indicators in Embodiment 1 of the present invention;

Fig. 7 is a control performance comparison diagram of Embodiment 2 of the present invention;

FIG. 8 is a comparison table of control performance indicators in Embodiment 2 of the present invention;

In the figure: 1: setting unit; 2: trend event-driven strategy unit; 3: fuzzy control unit.

Detailed ways

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

The present invention discloses a fuzzy control method for zinc smelting and roasting process driven by trend events. The structural block diagram of the control method is shown in Fig. 1, including the following steps:

S1. The setting unit 1 sets the temperature setting value of the roasting process and the control cycle N of the fuzzy controller.

S2. The trend event-driven strategy unit 2 calculates the temperature deviation and extracts the temperature trend according to the set value and the real-time sampling temperature value of the sensor, and the event-driven strategy based on the temperature trend screens out the event when the working condition changes or reaches the preset control cycle The temperature trend and temperature deviation are timely input to the fuzzy control unit 3 for fuzzy control.

The event-driven strategy based on temperature trend described in step S2, as shown in Figure 2, includes the following steps:

S2.1. Input the given control cycle N, the temperature data within the time window (t ₁ , t _i );

t ₁ and t _i are the starting point and end point of the time window respectively, and the control cycle N is the maximum length of the time window. In order to ensure the effect of fitting the model, a minimum time window length l _th must be set, that is, l _th ≤ i ≤ N, if i<l _th , wait for more temperature data until i≥l _th .

S2.2. Build a trend model;

For the temperature data within the time window (t ₁ , t _i ), use the least square method to establish a trend model:

Among them, Y _i =[y ₁ ,y ₂ ...y _i ] ^T , are the parameters of the model, is an estimate of the noise bias, is the output of the model, that is, the temperature prediction value, y _k is the temperature detection value, t _k is the time value, 1≤k≤i≤N, 1≤m≤3.

S2.3. Determine whether the temperature trend has changed. If not, go to step S2.4. If it has changed, that is, the trend event is triggered, go to step S2.6.

Determining whether the temperature trend has changed involves the following steps:

S2.3.1. Calculate prediction error e _i+1 and first-type threshold th _1,i :

in, Represents the t distribution, and α is the confidence level of the t distribution;

Judging whether |e _i+1 |≤th _1,i , if yes, proceed to step S2.3.4;

S2.3.2. If the result of S2.3.1 is no, then judge whether the current temperature y _i+1 is an abnormal value, that is, for the time window (t _i+1 , ) for all moments t _j , i+1≤j≤i+l _th , the corresponding forecast error e _j and the first-class threshold th _1,j-1 are calculated by the formula in S2.3.1,

And judge whether all |e _j |≥th _1,j-1 , if not, then y _i+1 is an abnormal value, and enter step S2.3.4;

S2.3.3. If the result of S2.3.2 is yes, the current temperature y _i+1 is not an abnormal value, and the temperature trend has changed;

S2.3.4. Calculate the cumulative error cusum(t _i+1 ) and the second-type threshold th _2,i :

cusum(t _i+1 )＝cusum(t _i )+e _i+1

b _i+1 ＝[(a _i +b _i ) ^T 1] ^T

Among them, _i≥lth ,

Determine whether |cusum(t _i+1 )|≤th _2,i , if yes, the temperature trend has not changed; if not, the temperature trend has changed.

S2.4. Judging whether i+1=N, that is, whether the length of the time window has reached the given control cycle, if the result is yes, the cycle event is triggered, and then proceed to step S2.6.

S2.5. If the result of S2.4 is negative, add the current temperature y _i+1 to the previous time window, and return to step S2.2.

S2.6. Calculate the first-order derivative of the trend model at the current moment; in order to measure the current trend, the first-order derivative of the trend model at the current moment is used as a measurement index, and its calculation formula is:

To generalize the method, the time window (t ₁ , _ti ) is normalized to [0,1]. At the end position of each trend, its first derivative The derivative value represents the trend of the previous temperature time series at the current moment, which can provide a more accurate reference for temperature control.

S2.7. Send the first-order derivative and temperature deviation at the current moment to the fuzzy controller. At this time, whether it is the triggering of the trend event in step S2.3 or the triggering of the periodic event in S2.4, it means that the working condition has changed. Therefore, the fuzzy controller will control the roasting process according to the temperature deviation and temperature trend.

S2.8. Determine whether to end the control, if yes, pause, if not, start a new time window from time t _i+1 , and go to step S2.2.

S3. The fuzzy control unit 3 is based on the rule setting based on expert experience, and fuzzifies the input temperature deviation and temperature trend. After fuzzy reasoning, the adjustment value of the feed amount is output after defuzzification, and the feed amount of the roasting process is controlled by the actuator. volume is adjusted.

In step S3, based on the rules of expert experience, three kinds of membership functions are set: Z-type membership function Z ^p (x), S-type membership function S ^p (x) and bell-type membership function They are:

Among them, a ^p , b ^p , c ^p , d ^p , and ν are the parameter values of the membership function, p=1, 2, 3 represent that the membership function belongs to the temperature deviation, temperature trend and feed amount respectively; q represents the different situations of the bell-shaped membership function.

The fuzzy description of the temperature deviation is: very high (VH), slightly high (LH), suitable (Z), slightly low (LL) and very low (VL). The S-type membership function is used to describe VH, the Z-type membership function is used for VL, and the bell-shaped membership function is used to describe the others.

Likewise, temperature trends are described fuzzily as: Very Large (HP), Slightly Larger (LP), Stable (S), Slightly Small (LN) and Very Small (HN). The S-type membership function is used to describe HP, the Z-type membership function is used for HN, and the bell-shaped membership function is used to describe others.

In order to achieve more precise control, the classification of the adjustment value of the feed amount is more detailed, and its fuzzy description is: positive big (PB), positive middle (PM), positive small (PS), zero (O), negative small (NS ), negative medium (NM) and negative large (NB). The S-type membership function is used to describe PB, the Z-type membership function is used for NB, and the bell-shaped membership function is used to describe others.

The detailed parameters of the membership function selection and description of the temperature deviation and temperature trend and the adjustment value of the feed amount are shown in Figure 3.

The fuzzy inference rules between the temperature deviation and the temperature trend and the adjustment value of the feed amount are given in Figure 4, and the sentence description is as follows:

If TD＝VH and TT＝HP or LP or S then U＝NB;

If TD＝VH and TT＝LN or HN then U＝NM;

If TD=LH and TT=HP or LP then U=NM;

If TD＝LH and TT＝S or LN or HN then U＝NS;

If TD=Z and TT=HP then U=NS;

If TD=Z and TT=LP or S or LN then U=O;

If TD=Z and TT=HN then U=PS;

If TD=LL and TT=HP or LP or S then U=PS;

If TD=LL and TT=LN or HN then U=PM;

If TD=VL and TT=HP or LP then U=PM;

If TD＝VL and TT＝S or LN or HN then U＝PB;

Among them, TD is the temperature deviation of fuzzification, TT is the temperature trend of fuzzification, and U is the adjustment value of feed amount of fuzzification.

In this method, the defuzzification method used is the center of gravity method, and its expression is:

Among them, u ^* is the adjustment value of the feed amount, that is, the output value of the fuzzy control, u _f is the adjustment value of the feed amount corresponding to the membership function, n is the number of the membership function of the adjustment value of the feed amount, Here, n=7.

Embodiment one:

In order to prove the effectiveness of this method, under the same initial conditions, the temperature setting value of the roasting process is set to 910°C, and the performance of the proposed fuzzy control method and conventional fuzzy control method is compared. The conventional fuzzy control method has the same membership function and fuzzy inference rules as the proposed method, but the difference is that the conventional method uses the rate of change of temperature deviation and has no corresponding event-driven strategy.

The control effect comparison is shown in Figure 5. The overshoot of the fuzzy control method proposed in this method is 0.2706, and the adjustment time is 23 minutes, while the overshoot of the conventional fuzzy control method is 0.4829, and the adjustment time is 81 minutes. Compared with the conventional fuzzy control method, the proposed fuzzy control method has smaller overshoot and adjustment time. The specific comparison is shown in Figure 6.

Embodiment two:

In order to prove that the proposed method can effectively deal with the change of working conditions, when the set value is 910°C and both controllers reach a steady state, a step signal with an amplitude of 10°C is added to the system to simulate the working conditions change.

The control effect is shown in Figure 7. The overshoot of the fuzzy control method proposed in this method is 1.0573, and the adjustment time is 57 minutes, while the overshoot of the conventional fuzzy control method is 1.0147, and the adjustment time is 207 minutes. Compared with the conventional fuzzy control method, the proposed fuzzy control method has a shorter adjustment time and no oscillation in the adjustment process. Since the proposed method has a smaller steady-state deviation, the overshoot of the system will be larger after the step signal is added. The specific comparison of the control performance after the operating condition transition is shown in Fig. 8.

In summary, through the above-mentioned fuzzy control method of zinc smelting and roasting process based on trend event drive, it has the following advantages:

(1) The present invention obtains the temperature trend in real time, utilizes the trend-driven strategy based on the temperature trend, and performs fuzzy control in time according to the temperature trend and temperature deviation when the working condition changes or reaches the preset control cycle, and then controls the feed amount Make adjustments to stabilize the system temperature, enrich the sampling data, and filter the input data of fuzzy control to make the whole control process more timely, more accurate and the system more stable;

(2) The trend-driven strategy based on the temperature trend of the present invention considers the temperature trend change, the temperature trend does not change and the temperature gradually shifts two situations, the working condition evaluation is more accurate, and the fuzzy control is adjusted to deal with the change of the working condition. The influence of the system makes the control performance of fuzzy control better;

(3) The input temperature deviation and temperature trend of the fuzzy control of the present invention, rather than the increment of temperature deviation and deviation, conform to the dynamic characteristics of the roasting process, have smaller overshoot, and the adjustment time is faster. When changing, the adjustment time is faster, the temperature control is more stable, and it is less affected by the change of working conditions;

(4) The present invention judges the abnormal value, which reduces the possibility of misjudgment;

(5) The present invention uses fuzzy control, which makes up for the deficiency of the mathematical model to a certain extent, and the defuzzification of the method adopts the center of gravity method, which is more intuitive and reasonable.

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 them; although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art can Various modifications and variations are made within the scope and scope of the invention, and such modifications and variations fall within the scope defined by the appended claims.

Claims (10)

1. A fuzzy control method for a zinc smelting roasting process based on trend event driving is characterized by comprising the following steps:

s1, setting a temperature set value and a control period N;

s2, calculating temperature deviation and extracting temperature trend according to a set value and a real-time sampling temperature value of a sensor, screening out the temperature trend and the temperature deviation when the working condition changes or reaches a preset control period based on an event-driven strategy of the temperature trend, and timely inputting the temperature trend and the temperature deviation to a fuzzy control unit;

and S3, setting rules based on expert experience, fuzzifying the input temperature deviation and temperature trend, performing fuzzy reasoning, performing defuzzification, outputting an adjustment value of the feeding amount, and adjusting the feeding amount in the roasting process by an actuator.

2. The fuzzy control method for the zinc smelting roasting process based on trend event driving as claimed in claim 1, wherein the event driving strategy based on temperature trend in step S2 comprises the following steps:

s2.1, inputting a given control period N and a time window (t)₁,t_i) Temperature data of the inside;

s2.2, constructing a trend model;

s2.3, judging whether the temperature trend changes, if not, entering a step S2.4, and if so, entering a step S2.6;

s2.4, determining whether i +1 is equal to N, if so, proceeding to step S2.6;

s2.5, if the result of the S2.4 is negative, returning to the step S2.2;

s2.6, calculating a first derivative of the trend model at the current moment;

s2.7, sending the first derivative and the temperature deviation of the current moment to a fuzzy controller;

s2.8, judging whether to finish control, if so, pausing, otherwise, stopping from t_i+1A new time window is started from time to time and the process proceeds to step S2.2.

3. The fuzzy control method for the zinc smelting roasting process based on trend event driving as claimed in claim 2, wherein the step of judging whether the temperature trend changes or not in step S2.3 comprises the following steps:

s2.3.1 calculating the prediction error e_i+1And a first class threshold th_1,iDetermine whether | e_i+1|≤th_1,iIf yes, go to step S2.3.4;

s2.3.2, if the result of S2.3.1 is negative, then the current temperature y is judged_i+1Whether or not it is an outlier, i.e. for a time windowAll time instants t in_j，i+1≤j≤i+l_thCalculating its corresponding prediction error e_jAnd a first class threshold th_1,j-1And determining whether all | e_j|≥th_1,j-1If not, then y_i+1Is an outlier, proceed to step S2.3.4;

s2.3.3, if S2.3.2 results in yes, then the current temperature y_i+1Not an outlier, the temperature trend has changed;

s2.3.4, calculating the cumulative error cusum (t)_i+1) And a threshold th of the second kind_2,i. Judging whether | cusum (t)_i+1)|≤th_2,iIf yes, the temperature trend is not changed, and if no, the temperature trend is changed.

4. A trend event driven fuzzy control method for zinc smelting roasting process according to claim 2, characterized by the time window (t) in step S2.1₁,t_i) Has a length of i, l_thI is not less than i and not more than N, the control period N is the maximum length of the time window l_thIs the minimum time window length.

5. The fuzzy control method for the zinc smelting roasting process based on trend event driving as claimed in claim 2, wherein the trend model in step S2.2 is:

wherein, Y_i＝[y₁,y₂…y_i]^T， Are the parameters of the model and are,in order to estimate the deviation of the noise,as output of the model, i.e. temperature prediction, y_kAs measured temperature value, t_kIs a time value, k is more than or equal to 1 and less than or equal to i and less than or equal to N, and m is more than or equal to 1 and less than or equal to 3.

6. The fuzzy control method for the zinc smelting roasting process based on trend event driving as claimed in claim 3 or 4, wherein the prediction error e is_i+1Threshold th of the first class_1,iCumulative error cusum (t)_i+1) And a threshold th of the second kind_2,iRespectively as follows:

cusum(t_i+1)＝cusum(t_i)+e_i+1

a_i＝((T_i ^TT_i)^-1(T_i)^T)^T[1 (t_i+1-t₁) (t_i+1-t₁)²]^T

b_i+1＝[(a_i+b_i)^T 1]^T

wherein,representing the t distribution, alpha is the confidence level of the t distribution,

7. the fuzzy control method for the zinc smelting roasting process based on trend event driving as claimed in claim 2, wherein the first derivative of the current trend model in step S2.6 is:

8. the fuzzy control method for the zinc smelting roasting process based on trend event driving as claimed in claim 1, wherein the fuzzification in the step S3 is described as follows:

the fuzzification of the temperature deviation is described as: very high VH, slightly high LH, suitably Z, slightly low LL and very low VL; describing VH by adopting an S-type membership function, describing VL by using a Z-type membership function, and describing the rest by using a bell-shaped membership function;

the blurring of the temperature trend is described as: very large HP, slightly larger LP, stable S, slightly smaller LN, and very small HN; describing HP by adopting an S-type membership function, HN by using a Z-type membership function, and describing the others by using bell-shaped membership functions;

the fuzzification of the adjustment values for the feed rates is described as: positive big PB, positive PM, positive small PS, zero O, negative small NS, negative middle NM and negative big NB; and the PB is described by adopting an S-type membership function, the NB is described by a Z-type membership function, and the others are described by bell-shaped membership functions.

9. The trend event driven fuzzy control method for zinc smelting roasting process according to claim 8, wherein said Z-type membership function Z^p(x) S type membership function S^p(x) Sum-bell membership functionRule setting based on expert experience is as follows:

wherein, a^p、b^p、c^p、d^p、V and v are parameter values of a membership function, and p is 1,2 and 3; q ═ 2, …,2, p ═ 1,2,3, denotes the membership function for temperature deviation, temperature trend and feed rate, respectively; q represents different cases of bell-shaped functions.

10. The fuzzy control method for the zinc smelting and roasting process based on trend event driving as claimed in claim 1, wherein the anti-fuzzy method in step S3 is a gravity center method, and the expression is:

wherein u is^*Is a feed materialAdjustment value of quantity, i.e. output value of fuzzy control, u_fThe adjustment value of the feeding amount corresponding to the membership function is obtained, n is the number of the membership functions of the adjustment value of the feeding amount, and n is 7.