CN203933038U - From the grid-connected mixing photovoltaic power generation control system of net - Google Patents

Abstract

The utility model relates to an off-grid and grid-connected hybrid photovoltaic power generation control system. The control system is composed of photovoltaic array, grid-connected inverter, battery pack, bidirectional inverter, load, power meter, public grid, demand response control system and switch group. The photovoltaic array is connected to the AC side through a grid-connected inverter, and the battery pack is connected to the AC side through a bidirectional inverter; S1-S4) are connected; the entire off-grid and grid-connected hybrid power generation system is uniformly managed and controlled by the demand response control system. Through the opening and closing combination of the switch group, the system supports 8 different operating modes to achieve the optimization of economic benefits and the peak-shaving and valley-filling of the power grid, and extend the service life of the battery by limiting the discharge depth and charging and discharging power of the battery.

Description

Off-grid and grid-connected hybrid photovoltaic power generation control system

technical field

The utility model relates to the field of new solar energy power generation and application, in particular to an off-grid and grid-connected hybrid photovoltaic power generation control system.

Background technique

With the intensification of the energy crisis, while developing and utilizing renewable energy, how to use energy more reasonably has gradually become a social concern. In recent years, the growth rate of electric energy load is faster than that of electric quantity, which leads to the decrease of load rate of power grid and the increase of peak-to-valley difference. For power users, this will greatly increase the cost of electricity consumption, which is not conducive to economic benefits, and for public power grids, this will affect the reliability and stability of power grid operation. Therefore, it is of great significance to comprehensively utilize photovoltaic power generation and grid power, and to optimize the dispatching of distributed power systems based on peak and valley electricity prices.

Most of the current distributed power systems, although they can realize grid-connected and off-grid operation of the system, and have energy storage devices to improve energy utilization, but still have the following deficiencies:

1. The existing system does not optimize the economic operation of the system for peak and valley electricity prices. Although individual systems can support peak and valley electricity prices, the scheduling schemes of these systems are relatively simple, and the mode switching is only performed according to a few thresholds or conditions. Although this can reduce operating costs to a certain extent, it has not been optimized from the perspective of global optimality control, so the optimization effect is limited.

2. Existing systems usually lack the forecast of photovoltaic power generation at each time period. Although some systems involve power generation forecasting, the forecasting algorithm is relatively simple and the reliability of the forecast value is low. Without the forecast of power generation in each time period, it is impossible to arrange the scheduling plan of the day through the power generation, and it is difficult to achieve economic optimization.

3. The existing system does not consider the service life of the battery. During operation, the discharge depth of the battery cannot be controlled, and deep discharge will greatly reduce the service life of the battery.

Utility model content

In order to solve the deficiencies in the existing technology, the utility model provides an off-grid hybrid photovoltaic power generation control system that optimizes and controls the micro-grid composed of photovoltaic power generation devices, battery packs and public grids according to the time-of-use electricity price .

In order to achieve the above object, the utility model adopts the following technical solutions:

An off-grid and grid-connected hybrid photovoltaic power generation control system, which consists of photovoltaic arrays, grid-connected inverters, battery packs, bidirectional inverters, loads, power meters, public grids, demand response controllers, management computers and switch groups , the switch group includes switches S1-S4; the photovoltaic array is connected to the AC side through the grid-connected inverter, and the battery pack is connected to the AC side through the bidirectional inverter; the grid-connected inverter is connected to the grid through the switch S1, and the S3 is connected to the bidirectional inverter and connected to the load through switches S3 and S4; the bidirectional inverter is connected to the load through switches S4 and connected to the power grid through switches S4 and S2; the power grid is connected to the load through switch S2; the whole off-grid is connected to the grid The hybrid power generation system is uniformly managed and controlled by the demand response controller and the management computer; through different on-off combinations of switches S1-S4, the system supports shutdown mode, off-grid battery power supply mode, off-grid photovoltaic-battery working mode, and grid-only power supply mode , Grid power supply - charging mode, photovoltaic power supply - grid connection - charging mode, grid power supply - photovoltaic grid connection - charging mode and grid power supply - photovoltaic grid connection - discharge mode switching.

The management computer is equipped with a historical information database, based on the historical data of photovoltaic power generation and power load and local weather information, predicts the photovoltaic power generation curve and power load curve of the day, and sends the forecast results to the demand response through Ethernet controller.

The demand response controller is composed of a microcontroller, an Ethernet communication interface and an AC contactor control interface, receives the power prediction value from the management computer, optimizes scheduling based on this, and finally controls the switch group through the opening and closing of each switch. Close, adjust the working mode, and realize the scheduling plan.

The various operating modes described:

1) Stop mode, all switches are off;

2) Off-grid battery power supply mode, S1, S2, S3 are disconnected, S4 is closed, the load is powered by the battery alone, and the energy flows from the battery to the load;

3) Off-grid photovoltaic-battery working mode, which is divided into two situations due to the different charging and discharging states of the battery: when the switches S1 and S2 are off, and S3 and S4 are closed, when the power of the photovoltaic array is not enough to meet the load demand, Supplemented by the battery to supply power to the load together, the energy flows from the photovoltaic array and the battery to the load;

The switches S1 and S2 are disconnected, and S3 and S4 are closed. When the power of the photovoltaic array is sufficient to meet the demand of the load, the battery is charged, and the energy flows from the photovoltaic array to the battery and the load;

4) Power grid alone power supply mode, switches S1, S3, S4 are open, S2 is closed, the public grid alone supplies power to the load, and the energy flows from the public grid to the load;

5) Grid power supply-charging mode, S1 and S3 are disconnected, S2 and S4 are closed, the public grid supplies power to the load and charges the battery, and the energy flows from the public grid to the battery and the load;

6) Photovoltaic power supply - grid-connected - charging mode, switch S2 is open, S1, S3, S4 are closed, the power of the photovoltaic array is sufficient to meet the load demand, after the battery is charged, the excess power is sent to the public grid, and the energy flows from the photovoltaic array to the load , battery and public grid;

7) Grid power supply-photovoltaic grid-connected-charging mode, S3 is disconnected, S1, S2, and S4 are closed, on the one hand, the photovoltaic array outputs power to the public grid, and on the other hand, the public grid supplies power to the load and charges the battery;

8) Grid power supply - photovoltaic grid connection - discharge mode, S4 is open, S1, S2, S3 are closed, the photovoltaic array and battery output power to the public grid, and the public grid supplies power to the load.

An optimal control method for an off-grid and grid-connected hybrid photovoltaic power generation control system. According to the historical data of photovoltaic power generation and power load and weather information, the photovoltaic power generation curve and power load curve of the day are predicted, and according to the predicted photovoltaic power generation With the power load, the optimization goal is to minimize the operating cost, the charging and discharging power of the battery per hour, the exchange power with the grid and the operating mode of the system are the optimized variables, and the power balance condition and the upper and lower limits of the exchange power with the grid , Battery charging and discharging power limitation and state of charge limitation are constraints, and the scheduling decision is made through the particle swarm optimization algorithm to make decision scheduling on the system operation mode, so that the system operation cost is the lowest, and by limiting the battery discharge depth and charging and discharging power Extend its service life.

The photovoltaic output power prediction is: under various weather conditions, within a research period, the daily photovoltaic output power is evenly divided into several intervals from 0 to the maximum value, and the power in the same interval is regarded as a state; the day is divided into For multiple time periods, one time period is at least 1 hour, count the transfer times of photovoltaic output power in each time period in the research period, and obtain the state transition matrix corresponding to the time period; , find the statistical data corresponding to the weather, and then use the Markov chain method to predict the photovoltaic output power throughout the day.

Electric load prediction is as follows: within a research period, the daily electricity load is counted, and the output power is evenly divided into several intervals from 0 to the maximum value, and the power in the same interval is regarded as a state; the day is divided into multiple time periods , a time period of at least 1 hour, the number of power load transfers in each time period within the research period is counted, and the state transition matrix corresponding to the time period is obtained; after the system is officially operated, according to the statistical data, through the Markov chain method, Forecast electricity load throughout the day.

When performing power prediction, first obtain the initial state probability mass function of the first unit time, that is, the initial distribution p ₁ , let the state probability corresponding to the measured power at the initial moment be 1, and the other state probabilities be 0, and then use the state transition matrix to calculate The distribution probability of each state at any time, the formula is as follows:

p _m+1 = p _m P _m

Among them, p _m and p _m+1 represent the distribution probability row vectors of the mth time period and the m+1th time period respectively, P _m is the state transition matrix corresponding to each mth time period, and the m+1th time period is obtained After the distribution probability p _m+1 of the distribution, use the mathematical expectation method to get the predicted value at the m+1th moment, the calculation formula is as follows:

F _m+1 = p _m+1 P _EXP

Among them, P _EXP is the mathematical expectation matrix, F _m+1 is the power prediction value of the m+1th time period, after obtaining F _m+1 , repeat the above process until the power prediction value of all time periods in the whole day is obtained, and then The output power is sent to the demand response controller via Ethernet.

The processing flow of the decision algorithm is:

(1) Obtain the pre-set system operating parameters, cost parameters, electricity price parameters, photovoltaic power and electricity load;

(2) Obtain the preset particle swarm algorithm parameters, mainly including population size, maximum number of iterations, learning factor and inertia weight coefficient, and set the confidence level of the objective function and constraints containing random variables;

where the objective function is:

$\left\{\begin{array}{c}\mathrm{<span\; class="notranslate">\; Min</span>}\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; f</span>}}\\ \mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; .</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; .</span>\mathrm{<span\; class="notranslate">\; PR</span>}<span\; class="notranslate">\; \{</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; no</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; \&le;</span>\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; f</span>}}<span\; class="notranslate">\; \}</span><span\; class="notranslate">\; \&Greater\; Equal;</span>\mathrm{<span\; class="notranslate">\; \&beta;</span>}\end{array}\right.$

is the objective function

β is the confidence level given by

C _i is the operating cost of the i-th period, where C _i =T[J _buy,i P _buy,i +P _pv,i C _{pv_m} +|P _bt,i |C _{bt_m} -J _sel,i P _sel,i ]

T is the time interval of the unit period

n is the total number of time slots in the scheduling cycle

_Pbuy,i is the electric power purchased from the public grid in the i-th period

P _sel,i is the electric power output to the public grid in the i-th period

P _pv,i is the generating power of the photovoltaic array in the i-th period

P _bt,i is the charging and discharging power of the battery in the i-th period, the discharge is positive, and the charge is negative

J _buy,i is the price of electricity purchased from the public grid in the i-th period

J _sel,i is the price of electricity sold to the public grid in the i-th period

C _{pv_m} is the unit operating cost of the photovoltaic array

C _{bt_m} is the maintenance cost of the storage battery;

(3) Initialize the population, randomly generate the charging and discharging power of the battery in each scheduling period and the power purchased from the public grid to form a particle, and use the constraints to test the feasibility of the particle until all the particles are initialized; at the same time, randomly generate each particle the initial speed of

(4) Calculate the fitness value of each particle, and compare the fitness value of each particle with the individual extreme value. If the former is better, update the current particle individual extreme value and individual optimal position; otherwise, keep it unchanged;

(5) Compare all current individual extremums and global extremums, and take the best one to update the current global extremum and its global optimal position;

(6) Update the velocity and position of each particle, and check the feasibility of the particles through the constraints until all particles are feasible;

(7) Repeat (4)-(8) until the termination condition is met;

(8) Output the optimal solution, that is, the operating mode of the system at each hour, the charging and discharging power of the battery, and the power exchanged with the grid.

In the step (3), the constraints include electric power balance constraints, power exchange constraints with the grid, and battery constraints;

The electric power balance constraint formula is:

${\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; buy</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; PV</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; \&eta;</span>}}_{\mathrm{<span\; class="notranslate">\; inv</span>}}<span\; class="notranslate">\; +</span>\frac{{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; bt</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}}{{\mathrm{<span\; class="notranslate">\; \&eta;</span>}}_{\mathrm{<span\; class="notranslate">\; ch</span>}}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; sel</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; ld</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}$

P _buy,i +P _pv,i η _inv +P _bt,i η _dis -P _sel,i -P _ld,i ＝0

In the formula, P _pv,i is the photovoltaic power generation power in the i-th period, P _buy,i and P _sel,i are the power bought and sold from the grid in the i-th period, respectively, and P _bt,i is the power of the battery in the i-th period Charge and discharge power, P _ld,i is the load of the i-th period, η _inv is the efficiency of the inverter, η _ch is the charging efficiency of the storage battery, and η _dis is the discharge efficiency of the storage battery;

The formula for exchanging power with the grid is:

$\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; \&le;</span>{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; buy</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; \&le;</span>{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; buy</span>}}^{\mathrm{<span\; class="notranslate">\; max</span>}}$

$\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; \&le;</span>{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; sel</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; \&le;</span>{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; sel</span>}}^{\mathrm{<span\; class="notranslate">\; max</span>}}$

In the formula, are the maximum power values of electricity purchase and sale respectively;

The battery constraint formula is:

${\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; cbt</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}^{\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; \&le;</span>{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; bt</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; \&le;</span>{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; abt</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}^{\mathrm{<span\; class="notranslate">\; max</span>}}$

SOC _min ≤ SOC _i ≤ SOC _max

$\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; no</span>}}{\text{<span class="notranslate"> \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; bt</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}$

in, Respectively, the maximum power of charge and discharge of the battery in the i-th period, SOC _i is the state of charge of the battery in the i-th period, SOC _min and SOC _max are the minimum and maximum values of the battery state of charge respectively, here assume the capacity of the battery Unchanged, which means that the storage energy of the battery at the initial moment and the end moment of the dispatching period are equal.

Beneficial effects of the utility model

1. Accurate power prediction: Existing systems usually lack the prediction of photovoltaic power generation. Although some systems involve power prediction, the algorithm is relatively simple. The power generation at each time period of a day is used as the predicted value of the power generation at each time period of the day. And the utility model adopts the predictive method of Markov chain, and this method judges the overall change trend of the state through the initial probability of things being in different states and the transition probability between each state, so as to realize the prediction to the future state, so the prediction result more reliable.

2. Economic operation optimization: The economic operation optimization scheme of the existing system is relatively simple. Generally, the mode switching is performed only according to several thresholds or conditions. However, the actual situation is ever-changing, and the system also contains the random variable of photovoltaic power. Therefore, this method The global optimum cannot be achieved, that is, the operating cost of the whole day cannot be minimized. However, the utility model adopts the particle swarm optimization algorithm, fully considers the impact of peak and valley electricity prices during optimization, and takes the lowest cost throughout the day as the objective function, so the optimization result is closer to the real optimization.

3. "Peak shaving and valley filling" of the power grid: the economic operation optimization of the utility model is realized by purchasing electricity from the grid or charging the storage battery during the valley period, selling electricity to the grid during the peak period or preferentially using the electric energy of the storage battery, thus realizing The "peak shaving and valley filling" of the power grid is conducive to the smooth and efficient operation of the power grid.

4. Prolong battery life: When the utility model solves the objective function through the particle swarm algorithm, the limiting conditions of battery charge and discharge power and battery state of charge can be added, so the maximum discharge depth of the battery can be limited, thereby prolonging the service life of the battery.

Description of drawings

Fig. 1 is a system structure block diagram of the present utility model;

Fig. 2 is a schematic diagram of the shutdown mode;

Figure 3 is a schematic diagram of an off-grid battery power supply mode;

Figure 4 is a schematic diagram of the discharge of the battery in the off-grid photovoltaic-battery working mode;

Figure 5 is a schematic diagram of the battery charging situation in the off-grid photovoltaic-battery working mode;

Fig. 6 is a schematic diagram of a power grid alone power supply mode;

Fig. 7 schematic diagram of grid power supply-charging mode;

Figure 8 is a schematic diagram of photovoltaic power supply-grid-connected-charging mode;

Figure 9 is a schematic diagram of grid power supply-photovoltaic grid-connected-charging mode;

Figure 10 is a schematic diagram of grid power supply-photovoltaic grid-connected-discharge mode;

Fig. 11 is a processing flowchart of the decision algorithm.

Detailed ways

In order to make the purpose, technical solution and advantages of the utility model clearer, the utility model will be further described in detail below in conjunction with the accompanying drawings and embodiments. The specific embodiments described here are only used to explain the utility model, and are not intended to limit the utility model.

Fig. 1 is a system structure diagram of the utility model. The utility model constructs a system for optimizing and controlling the micro-grid based on the time-of-use electricity price, which consists of photovoltaic arrays, grid-connected inverters, battery packs, bidirectional inverters, loads, power meters, AC busbars, public grids, Demand response controller, management computer and switch group constitute.

The demand response controller can control four switches S1-S4, and each switch has two states of closed and open. By changing the state of S1-S4, the system can work in 8 different operating modes:

1) Shutdown mode. All switches are off, and the entire system works in shutdown mode without energy flow, as shown in Figure 2.

2) Off-grid battery power supply mode. S1, S2, and S3 are disconnected, S4 is closed, the load is powered by the battery alone, and the energy flows from the battery to the load, as shown in Figure 3.

3) Off-grid photovoltaic-battery working mode. This mode is divided into two situations due to the different charging and discharging states of the battery: when S1 and S2 are disconnected, and S3 and S4 are closed, when the power of the photovoltaic array is insufficient to meet the load demand, the battery will supplement it to supply power to the load together, and the energy The photovoltaic array and battery flow to the load, as shown in Figure 4; S1 and S2 are disconnected, and S3 and S4 are closed. When the power of the photovoltaic array is sufficient to meet the load demand, the battery is charged, and the energy flows from the photovoltaic array to the battery and the load, such as Figure 5 shows.

4) Power grid alone power supply mode. S1, S3, and S4 are disconnected, S2 is closed, the public grid alone supplies power to the load, and energy flows from the public grid to the load, as shown in Figure 6.

5) Grid power supply - charging mode. S1 and S3 are disconnected, S2 and S4 are closed, the public grid supplies power to the load and charges the battery, and the energy flows from the public grid to the battery and the load, as shown in Figure 7.

6) Photovoltaic power supply - grid connection - charging mode. S2 is open, S1, S3, and S4 are closed, and the power of the photovoltaic array is sufficient to meet the load demand. After the battery is charged, the excess power is sent to the public grid. Energy flows from the PV array to the load, battery and public grid, as shown in Figure 8.

7) Grid power supply - photovoltaic grid connection - charging mode. S3 is open, and S1, S2, and S4 are closed. On the one hand, the photovoltaic array outputs power to the public grid, and on the other hand, the public grid supplies power to the load and charges the battery, as shown in Figure 9.

8) Grid power supply - photovoltaic grid connection - discharge mode. S4 is open, S1, S2, and S3 are closed, and the photovoltaic array and storage battery output power to the public grid, and the public grid supplies power to the load, as shown in Figure 10.

When the system is running, the management computer will use the Markov chain method to predict the photovoltaic output power and electricity load of the day. Therefore, before the official operation of the system, the historical data of photovoltaic output power and electric load need to be counted and stored in the historical information database. The specific statistical method of photovoltaic output power is:

1) Count the daily photovoltaic output power within a research period under various weather conditions.

2) The output power is evenly divided into several intervals from 0 to the maximum value, and the power in the same interval is regarded as a state.

3) Divide a day into multiple time periods (one time period is at least 1 hour), take 1 hour as the minimum time interval, count the transfer times of photovoltaic output power in each time period within the research period, and obtain the state transition corresponding to this time period matrix.

The statistical method of electricity load is basically the same as the statistical method of photovoltaic output power, except that the weather factor does not need to be considered in the statistics.

After the system is officially running, the power forecasting software on the management computer first obtains the weather information provided by the meteorological center, and finds the statistical data under the corresponding weather in the database, and then uses the Markov chain method to calculate the photovoltaic output power of the whole day. Make predictions.

When performing power prediction, the initial state probability mass function of the first unit time is first obtained, that is, the initial distribution p ₁ , and the state probability corresponding to the measured power at the initial moment is 1, and the other state probabilities are 0. Then use the state transition matrix to calculate the distribution probability of each state at any time, the formula is as follows:

p _m+1 = p _m P _m

Among them, p _m and p _m+1 represent the distribution probability row vectors of the m-th time period and the m+1-th time period respectively, and P _m is the state transition matrix corresponding to each m-th time period. After obtaining the distribution probability p _m+1 of the m+1th time period, the predicted value at the m+1th moment is obtained by using the mathematical expectation method. The calculation formula is as follows:

F _m+1 = p _m+1 P _EXP

Among them, P _EXP is the mathematical expectation matrix, and F _m+1 is the power prediction value of the m+1th time period.

After obtaining F _m+1 , the above process is repeated until the predicted power values of all periods of the day are obtained.

The output power is then sent to the demand response controller via Ethernet. According to the predicted photovoltaic power generation and power load, the demand response controller takes the minimum operating cost as the optimization goal, takes the charging and discharging power of the battery per hour, the exchange power with the grid and the operating mode of the system as the optimized variables, and uses the electric energy The balance condition, the upper and lower limits of the power exchanged with the grid, the battery charge and discharge power limit, and the state of charge limit are constrained conditions, and the scheduling decision is made through the particle swarm optimization algorithm. The processing flow of the decision algorithm is shown in Figure 11. The specific process is as follows:

(1) Obtain the pre-set system operating parameters, cost parameters, electricity price parameters, photovoltaic power and electricity load.

(2) Obtain the pre-set particle swarm optimization parameters, mainly including population size, maximum number of iterations, learning factors c ₁ , c ₂ and inertial weight coefficient ω, etc., and set the confidence level of the objective function and constraints containing random variables .

where the objective function is:

$\left\{\begin{array}{c}\mathrm{<span\; class="notranslate">\; Min</span>}\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; f</span>}}\\ \mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; .</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; .</span>\mathrm{<span\; class="notranslate">\; PR</span>}<span\; class="notranslate">\; \{</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; no</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; \&le;</span>\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; f</span>}}<span\; class="notranslate">\; \}</span><span\; class="notranslate">\; \&Greater\; Equal;</span>\mathrm{<span\; class="notranslate">\; \&beta;</span>}\end{array}\right.$

In the formula, is the objective function, β is a given confidence level, and C _i is the operating cost of the i-th period. Where C _i ＝T[J _buy,i P _buy,i +P _pv,i C _{pv_m} +|P _bt,i |C _{bt_m} -J _sel,i P _sel,i ], T is the time interval of a unit period, n is the total number of periods in the scheduling cycle, P _buy,i is the electric power purchased from the public grid in the i-th period, P _sel,i is the electric power output to the public grid in the i-th period, P _pv,i is the power generation of the photovoltaic array in the i-th period Power, P _bt,i is the charging and discharging power of the battery in the i-th period, the discharge is positive, and the charge is negative, J _buy,i is the price of electricity purchased from the public grid in the i-th period, J _sel,i is the electricity sold to the i-th period The electricity price of the public grid, C _{pv_m} is the unit operating cost of the photovoltaic array, and C _{bt_m} is the maintenance cost of the storage battery.

(3) Initialize the population. Randomly generate the charging and discharging power of the battery in each scheduling period and the power purchased from the public grid to form a particle, and use the constraints to test the feasibility of the particle until all the particles are initialized. At the same time, the initial velocity of each particle is randomly generated.

The constraints include electric power balance constraints, power exchange constraints with the grid, and storage battery constraints.

The electric power balance constraint formula is:

${\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; buy</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; PV</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; \&eta;</span>}}_{\mathrm{<span\; class="notranslate">\; inv</span>}}<span\; class="notranslate">\; +</span>\frac{{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; bt</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}}{{\mathrm{<span\; class="notranslate">\; \&eta;</span>}}_{\mathrm{<span\; class="notranslate">\; ch</span>}}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; sel</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; ld</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}$

P _buy,i +P _pv,i η _inv +P _bt,i η _dis -P _sel,i -P _ld,i ＝0

In the formula, P _pv,i is the photovoltaic power generation power in the i-th period, P _buy,i and P _sel,i are the power bought and sold from the grid in the i-th period, respectively, and P _bt,i is the power of the battery in the i-th period Charge and discharge power, P _ld,i is the load of the i-th period, η _inv is the efficiency of the inverter, η _ch is the charging efficiency of the storage battery, and η _dis is the discharge efficiency of the storage battery;

The formula for exchanging power with the grid is:

$\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; \&le;</span>{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; buy</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; \&le;</span>{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; buy</span>}}^{\mathrm{<span\; class="notranslate">\; max</span>}}$

$\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; \&le;</span>{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; sel</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; \&le;</span>{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; sel</span>}}^{\mathrm{<span\; class="notranslate">\; max</span>}}$

In the formula, are the maximum power values of electricity purchase and sale respectively.

The battery constraint formula is:

${\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; cbt</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}^{\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; \&le;</span>{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; bt</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; \&le;</span>{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; abt</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}^{\mathrm{<span\; class="notranslate">\; max</span>}}$

SOC _min ≤ SOC _i ≤ SOC _max

$\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; no</span>}}{\text{<span class="notranslate"> \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; bt</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}$

in, are the maximum charging and discharging power of the battery in the i-th period, SOC _i is the state of charge of the battery in the i-th period, SOC _min and SOC _max are the minimum and maximum values of the battery state of charge respectively, here it is assumed that the capacity of the battery is not Change, indicating that the storage energy of the storage battery at the initial moment and the end moment of the dispatching cycle is equal.

Due to the randomness of photovoltaic output power, the existence of this random variable makes certain constraints no longer deterministic. Therefore, the probability form is used to describe the inequality constraints containing random variables, so that they can be established under a certain confidence level, so as to realize the processing of these constraints.

(4) Calculate the fitness value of each particle, and compare the fitness value of each particle with the individual extreme value. If the former is better, update the current particle individual extreme value and individual optimal position; otherwise, keep it unchanged.

(5) Compare all the current individual extremums with the global extremum, and take the best one to update the current global extremum and its global optimal position.

(6) Update the velocity and position of each particle, and check the feasibility of the particles through the constraints until all the particles are feasible.

(7) Repeat (4)-(7) until the termination condition is met (the maximum number of iterations is reached, the best solution does not change for several consecutive times or the difference between the best solution and the average fitness value is less than a certain set constant).

(8) Output the optimal solution, that is, the operating mode of the system at each hour, the charging and discharging power of the battery, and the power exchanged with the grid.

After the decision is made, the demand response controller controls the on-off of the switch group according to the decision information, so that the system runs in the specified mode and the scheduling scheme has been implemented.

Claims (1)

1. The off-grid and grid-connected hybrid photovoltaic power generation control system is characterized by comprising a photovoltaic array and a storage battery pack, wherein the photovoltaic array is connected with a switch group through a grid-connected inverter, the storage battery pack is connected with the switch group through a bidirectional inverter, the grid-connected inverter is connected with a public network through a switch S1, the grid-connected inverter is connected with the bidirectional inverter through a switch S3, the bidirectional inverter is connected with a load through a switch S4, the bidirectional inverter is connected with the public network through switches S4 and S2 in sequence, the public network is connected with the load through a switch S2, the switch group is further connected with a demand response controller and a management computer in sequence, and the switch S1 and the switch S2 are connected with the public network through an ammeter respectively.