A method for optimizing energy storage configuration considering the SOC constraint of energy storage

Abstract

The present invention discloses an energy storage configuration optimization method considering the energy storage SOC constraint. According to the behavior indicators of the current photovoltaic energy storage power station, the direction of the current energy storage action is pre-judged; the actual energy storage action direction is set, that is, when the specified energy storage action power value is positive, the energy storage action is discharging; when the energy storage action value is negative, the energy storage action is discharging; the iteration number trigger probability function trigger_i and the simulation time trigger probability function trigger_h are set; the improved particle swarm algorithm is used to output the optimal solution of the energy storage configuration. The above method can find a better particle evolution direction, increase the number of excellent particles, thereby improving the calculation speed and solution efficiency of the algorithm.

Description

Technical Field

[0001] The present invention relates to the technical field of power energy storage, and particularly relates to an energy storage configuration optimization method considering the energy storage SOC constraint. Background Art

[0002] Power energy storage technology can improve the operation efficiency, safety, economy of a conventional power system and the utilization efficiency of renewable energy. The application of power energy storage technology involves all aspects of "source-load-grid", and it is an effective way to improve the operation efficiency and power supply quality of a conventional power system. In addition, with the increasing rise of distributed generation systems and microgrids, power energy storage technology can improve the power supply capacity of distributed generation systems and microgrids at a relatively low cost. On the other hand, except for pumped-storage energy storage, compressed air energy storage and lead-acid batteries, the costs of other power energy storage systems are still relatively high. Pumped-storage energy storage and lead-acid batteries are restricted by geographical locations and environmental factors, and there are certain bottlenecks in further development. Therefore, reasonably configuring the capacity and power of energy storage to reduce the configuration cost of the energy storage system has become the primary problem in the energy storage configuration work.

[0003] In terms of the solution method, relevant analyses mostly adopt intelligent optimization algorithms, such as particle swarm algorithm, genetic algorithm, bat algorithm, simulated annealing algorithm, etc. During specific implementation, first set the operation strategy according to the system characteristics, and then use the intelligent algorithm to perform iterative optimization repeatedly. The intelligent algorithm has good performance in the field of solving nonlinear problems, but the problem of slow solving speed is also very prominent, which has always been an urgent problem for relevant staff to solve.

[0004] Optimizing the evolution method of the individuals of the intelligent algorithm is an important way to improve the solving speed of the intelligent algorithm. When the current intelligent algorithm considers the energy storage SOC constraint, the commonly used method is to verify the charge and discharge power of the energy storage at the current moment with the SOC of the energy storage at the previous moment. If the limit is exceeded, the limit value is assigned to the charge and discharge power of the energy storage at the current moment. This method can solve the problem of the energy storage SOC constraint. However, because the verification is carried out in chronological order, the energy storage action value at the previous moment will affect the energy storage action value at the next moment; the value at the previous moment will not change after verification; if the energy storage action value at the previous moment satisfies the SOC verification, but the corresponding value is a poor solution, it will lead to the meaninglessness of the calculation at the subsequent moments of this individual, which greatly increases the calculation time of the algorithm.

[0005] It can be seen that in the current solving model of the intelligent algorithm considering the energy storage SOC constraint, the proportion of poor solutions caused by the SOC constraint increases, which slows down the convergence speed and affects the overall solving efficiency. Summary of the Invention

[0006] The objective of the present invention is to provide an energy storage configuration optimization method considering the energy storage SOC constraint. This method can find a better particle evolution direction, increase the number of excellent particles, thereby improving the calculation speed and solution efficiency of the algorithm.

[0007] The objective of the present invention is achieved through the following technical solutions:

[0008] An energy storage configuration optimization method considering the energy storage SOC constraint, the method comprising:

[0009] Step 1: According to the behavior indicators of the current photovoltaic energy storage power station, preliminarily judge the direction of the current energy storage action, that is, judge whether the current energy storage action is charging or discharging;

[0010] Step 2: Set the actual energy storage action direction, that is, it is stipulated that when the energy storage action power value is positive, the energy storage action is discharging; when the energy storage action value is negative, the energy storage action is charging;

[0011] Step 3: Set an iteration number trigger probability function trigger_i and a simulation time trigger probability function trigger_h; wherein, the iteration number trigger probability function trigger_i decreases from the maximum trigger probability trigger_i_max to the minimum trigger probability trigger_i_min as the number of iterations increases; the simulation time trigger probability function trigger_h decreases from the maximum trigger probability trigger_h_max to the minimum trigger probability trigger_h_min as the simulation time increases;

[0012] Step 4: Incorporate an evolution direction correction mechanism into the traditional particle swarm algorithm, and iteratively solve the optimal solution of the energy storage configuration.

[0013] It can be seen from the technical solutions provided by the present invention described above that the above method can find a better particle evolution direction, increase the number of excellent particles, thereby improving the calculation speed and solution efficiency of the algorithm. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0015] Figure 1 It is a schematic flowchart of the energy storage configuration optimization method considering the energy storage SOC constraint provided by the embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0016] The following clearly and completely describes the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments, which does not constitute a limitation to the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

[0017] As Figure 1 shown is a schematic flowchart of an energy storage configuration optimization method considering the energy storage SOC constraint provided by an embodiment of the present invention. The method includes:

[0018] Step 1: According to the behavior indicators of the current photovoltaic and energy storage power station, pre-judge the direction of the current energy storage action, that is, judge whether the current energy storage action is charging or discharging;

[0019] In this step, in the current photovoltaic and energy storage power station, the actual output of the photovoltaic needs to follow the planned output curve. When the actual output of the photovoltaic exceeds the planned output curve, the probability that the current energy storage charging behavior is the optimal solution is greater. When the actual output of the photovoltaic is lower than the planned output curve, the probability that the current energy storage discharging behavior is the optimal solution is greater, which is expressed as:

[0020]

[0021] In the formula, PV is the actual output of the photovoltaic; PV f is the planned output of the photovoltaic; E op-fc =-1 represents that the probability that the current energy storage action being charging is the optimal solution is greater; E op-fc =1 represents that the probability that the current energy storage action being discharging is the optimal solution is greater.

[0022] Step 2: Set the actual action direction of the energy storage, that is, it is stipulated that when the energy storage action power value is positive, the energy storage action is discharging; when the energy storage action value is negative, the energy storage action is charging;

[0023] In this step, the set actual action direction of the energy storage is expressed as:

[0024]

[0025] In the formula, P(t) is the random particle of the particle swarm algorithm and is also the energy storage action power value; E op-re =1 indicates that the energy storage action is discharging; E op-re =-1 indicates that the energy storage action is charging.

[0026] Step 3: Set the iteration number trigger probability function trigger_i and the simulation time trigger probability function trigger_h;

[0027] Among them, the iteration number trigger probability function trigger_i decreases from the maximum trigger probability trigger_i_max to the minimum trigger probability trigger_i_min as the iteration number increases; the simulation time trigger probability function trigger_h decreases from the maximum trigger probability trigger_h_max to the minimum trigger probability trigger_h_min as the simulation time increases.

[0028] In this step, the iteration number trigger probability function trigger_i is expressed as:

[0029]

[0030] In the formula, r1 is a random number with a value range of [-1, 1]. The selection of r1 is the random number required for the particle swarm algorithm, and no additional random number needs to be generated; maxIterations is the maximum number of iterations; i is the current iteration number.

[0031] Here, the iteration number trigger probability function trigger_i is set to prevent the algorithm from falling into a local optimum and to avoid artificial intervention from affecting the calculation accuracy of the algorithm. The algorithm is triggered with a certain probability. As the iteration number increases, the triggering probability decreases, gradually relaxing the restriction on the particle evolution direction. When the value of trigger_i is 1, it is the triggering probability function; when the value of trigger_i is -1, it is the non-triggering probability function.

[0032] The simulation time trigger probability function trigger_h is expressed as:

[0033]

[0034] In the formula, r2 is a random number with a value range of [-1, 1]. The selection of r2 is the random number required for the particle swarm algorithm, and no additional random number needs to be generated; h is the current energy storage action time.

[0035] Since in the existing SOC constraint process, the accuracy of the previous moment's value has a greater impact on the particle accuracy, the simulation time trigger probability function trigger_h is set. As the simulation time increases, the triggering probability of the algorithm decreases, reducing the influence of artificial intervention on the particle evolution direction and avoiding local optima. Similarly, when the value of trigger_h is 1, it is the triggering probability function; when the value of trigger_h is -1, it is the non-triggering probability function.

[0036] Among them, r1 and r2 cannot use the same random number to avoid coupling.

[0037] Step 4: Incorporate the evolution direction correction mechanism into the traditional particle swarm algorithm and iteratively solve the optimal solution of the energy storage configuration.

[0038] In this step, the specific process is as follows:

[0039] 1) First, input the data information required for energy storage configuration;

[0040] Including operating parameters: the actual output curve of photovoltaic power generation. In this embodiment, typical daily power generation data is adopted; and the planned

[0041] output curve. The specific operating parameters are shown in Table 1 below:

[0042] Table 1 Operating Parameters

[0043]

[0044] Energy storage system parameters: the maximum and minimum capacities of the energy storage system, the maximum and minimum powers of the energy storage system, the initial SOC,

[0045] the upper and lower limits of SOC and the energy conversion efficiency. The specific energy storage system parameters are shown in Table 2 below:

[0046] Table 2 Energy Storage System Parameters

[0047]

[0048]

[0049] 2) Then set the relevant parameters of the improved particle swarm optimization algorithm;

[0050] The specific parameters are shown in Table 3 below:

[0051] Table 3 Reference Values of Simulation Parameters

[0052]

[0053] 3) Initialize the energy storage action power value P(t) at each moment as the decision variable, initialize the evolution speed V(t), and initialize the fitness function planned output deviation degree, which is expressed as:

[0054]

[0055] The sum of the photovoltaic output and the energy storage output is the actual output of the photovoltaic power station. The objective function is expressed as the minimum deviation between the actual output of the photovoltaic power station and the planned output.

[0056] 4) According to the particle swarm optimization algorithm, perform speed update:

[0057] V(i + 1) = ω·V(i) + c1r1·(P b (i) - X(i)) + c2r2·(P gd - X(i)) (6)

[0058] Wherein, V(i) is the particle evolution speed; P b is the individual optimal particle; P gb is the global optimal particle; c1 and c2 are acceleration constants; ω is the inertia constant; X is the random particle generated by the particle swarm.

[0059] 5) According to the input data information, judge whether the energy storage action should discharge or charge at each moment, that is, when E op-fc equals 1, the discharge probability should be greater, and when E op-fc equals -1, the charge probability should be greater. For example, in a microgrid, when the photovoltaic output is less than the load demand, the probability of the energy storage discharging is greater at this time, and it is judged that the energy storage action is discharging at this time; there are many targets that can be selected in practice, such as the energy storage operating in the peak-valley arbitrage mode. When the electricity price is high, the probability of discharging is high, and it is judged that the energy storage action is discharging at this time. In this embodiment, the planned output deviation degree is selected;

[0060] 6) Calculate the iteration number trigger probability function trigger_i set in step 3. The iteration number trigger probability function trigger_i decreases from trigger_i_max to trigger_i_min as the iteration number increases; because two random numbers need to be calculated when updating the particle swarm algorithm speed, when calculating the random number required for this iteration number trigger probability function trigger_i, use a random number of the improved particle swarm algorithm speed update function to reduce the operation time;

[0061] 7) Calculate the simulation time trigger probability function trigger_h set in step 3. The simulation time trigger probability function trigger_h decreases from trigger_h_max to trigger_h_min as the simulation time t increases; when calculating the random number required for this simulation time trigger probability function trigger_h, use a random number of the improved particle swarm algorithm speed update function to reduce the operation time; however, note that in order to avoid the coupling of trigger_h and trigger_i, do not select the same random number;

[0062] 8) Judge whether the energy storage action power value P(t) is the same as the desired energy storage action power calculated by the E op-fc index. If they are the same, update the particle position; if they are different, make the next judgment, that is

[0063]

[0064] If E op-fc ×E op-re = 1, it means that the energy storage action power value P(t) is the same as the desired energy storage action power calculated by the E op-fc index; if Eop-fc ×E op-re = -1, it indicates that the energy storage action power value P(t) is different from the desired energy storage action power calculated by the E op-fc index.

[0065] 9) If the energy storage action power value P(t) is different from the desired energy storage action power calculated by the E op-fc index, then further determine whether the direction of the particle evolution speed V(t) is towards the currently pre-judged energy storage action direction in step 1. If so, update the particle position; if not, proceed to the next judgment;

[0066] In specific implementation, since E op-fc ×E op-re = -1, if E op-fc = 1, at this time, it is desired that the energy storage is charged, then E op-re = -1, the energy storage is actually discharging, P(t)>0, at this time V(t)<0, making the probability of this particle evolving towards the optimal solution greater; if E op-fc = -1, at this time, it is desired that the energy storage discharges, then E op-re = 1, the energy storage is actually charging, P(t)<0, at this time V(t)>0, making the probability of this particle evolving towards the optimal solution greater.

[0067] 10) If it does not evolve in the same direction, then further judge the calculated iteration number trigger probability function trigger_i. If trigger_i is less than zero, it is not triggered, and the particle position is directly updated; if trigger_i is greater than zero, proceed to the next judgment;

[0068] 11) If trigger_i is greater than zero, then further judge the calculated simulation time trigger probability function trigger_h. If trigger_h is less than zero, it is not triggered, and the particle position is directly updated; if trigger_h is greater than zero, trigger the evolution direction correction function, reverse the velocity direction, so that the particle evolution direction is towards the direction where the energy storage power evolution direction is the same as the currently pre-judged energy storage action direction in step 1, that is:

[0069] V(t) = -V(t) (8)

[0070] 12) After correcting the particle evolution direction, update the particle position, that is:

[0071] X(i + 1) = X(i) + V(i + 1) (9)

[0072] 13) After updating the particle position, re-perform the operation in step 3) to calculate the fitness function;

[0073] 14) Determine whether the current iteration number i is greater than the iteration number value set by the simulation parameters. If the iteration number is not reached, return to step 5); if the iteration number is reached, output the optimal solution of the energy storage configuration and end the operation.

[0074] It should be noted that the content not described in detail in the embodiments of the present invention belongs to the prior art well-known to those skilled in the art.

[0075] In summary, the method described in the embodiments of the present invention can find a better particle evolution direction through the parameters in the model, increase the number of excellent particles, and thus improve the calculation speed of the algorithm; at the same time, as the iteration number increases or as the simulation step length decreases, the improved algorithm is more obvious in shortening the solution time and improves the solution efficiency.

[0076] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed by the present invention should be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims. The information disclosed in the background art part of this article is only intended to deepen the understanding of the overall background technology of the present invention, and should not be regarded as an admission or any form of suggestion that this information constitutes the prior art already known to those skilled in the art.

Claims

1. An energy storage configuration optimization method considering the SOC constraint of energy storage, characterized in that The method includes the following steps: Step 1: Based on the behavior indicators of the current photovoltaic and energy storage power station, pre-judge the direction of the current energy storage action, that is, judge whether the current energy storage action is charging or discharging; Step 2: Set the actual energy storage action direction, that is, it is stipulated that when the energy storage action power value is positive, the energy storage action is discharging; when the energy storage action value is negative, the energy storage action is charging; Step 3: Set the iteration number trigger probability function trigger_i and the simulation time trigger probability function trigger_h; among them, the iteration number trigger probability function trigger_i decreases from the maximum trigger probability trigger_i_max to the minimum trigger probability trigger_i_min as the number of iterations increases; the simulation time trigger probability function trigger_h decreases from the maximum trigger probability trigger_h_max to the minimum trigger probability trigger_h_min as the simulation time increases; In Step 3, the iteration number trigger probability function trigger_i is expressed as: In the formula, r1 is a random number with a value range of [-1, 1]. The selection of r1 is the random number required for the particle swarm algorithm and no additional random number needs to be generated; maxIterations is the maximum number of iterations; i is the current number of iterations; when the value of trigger_i is 1, it is the trigger probability function; when the value of trigger_i is -1, it is the non-trigger probability function; The simulation time trigger probability function trigger_h is expressed as: In the formula, r2 is a random number with a value range of [-1, 1]. The selection of r2 is the random number required for the particle swarm algorithm and no additional random number needs to be generated; h is the current energy storage action time; when the value of trigger_h is 1, it is the trigger probability function; when the value of trigger_h is -1, it is the non-trigger probability function; Among them, r1 and r2 cannot use the same random number to avoid coupling; Step 4: Add an evolution direction correction mechanism to the traditional particle swarm algorithm to iteratively solve the optimal solution of the energy storage configuration; The process of Step 4 is specifically as follows: 1) First, input the data information required for the energy storage configuration; 2) Then set the relevant parameters of the improved particle swarm algorithm; 3) Initialize the energy storage action power value P(t) at each moment as the decision variable, initialize the evolution speed V(t), and initialize the fitness function planned output deviation degree, which is expressed as: Among them, PV is the actual output of the photovoltaic; PV f is the planned output of the photovoltaic; P(t) is the random particle of the particle swarm algorithm and also the energy storage action power value; The sum of the photovoltaic output and the energy storage output is the actual output of the photovoltaic power station. The objective function is expressed as the minimum deviation between the actual output and the planned output of the photovoltaic power station; 4) Update the speed according to the particle swarm algorithm: V(i + 1)= ω·V(i)+ c1r1·(P b (i)-X(i))+ c2r2·(P gd -X(i)) (6) Where, V(i) is the particle evolution speed; P b is the individual optimal particle; P gb is the global optimal particle; c1 and c2 are acceleration constants; ω is the inertia constant; X is the random particle generated by the particle swarm; 5) Determine whether the energy storage should discharge or charge at each moment according to the input data information, i.e., E op-fc When it is equal to 1, the probability of discharging should be greater, E op-fc When it is equal to -1, the probability of charging should be greater; 6) Calculate the iteration number trigger probability function trigger_i set in Step 3. The trigger probability of this iteration number trigger probability function trigger_i decreases from trigger_i_max to trigger_i_min as the number of iterations increases; among them, when calculating the random number required for this iteration number trigger probability function trigger_i, use a random number of the improved particle swarm algorithm speed update function to reduce the operation time; 7) Calculate the simulation time trigger probability function trigger_h set in step 3. The trigger probability of this simulation time trigger probability function trigger_h decreases from trigger_h_max to trigger_h_min as the simulation time t increases. Among them, when calculating the random number required for this simulation time trigger probability function trigger_h, use a random number of the improved particle swarm algorithm velocity update function to reduce the operation time; 8) Determine whether the energy storage operation power value P(t) is the same as the desired energy storage operation power calculated by the E op-fc index. If they are the same, update the particle position; if they are different, proceed to the next judgment, that is If E op-fc ×E op-re = 1, it indicates that the energy storage action power value P(t) is the same as the desired energy storage action power calculated by the E op-fc index; if E op-fc ×E op-re = -1, it indicates that the energy storage action power value P(t) is different from the desired energy storage action power calculated by the E op-fc index; 9) If the energy storage operation power value P(t) is different from the desired energy storage operation power calculated by the E op-fc index, then determine again whether the direction of the particle evolution speed V(t) is towards the current energy storage operation direction pre-determined in step 1. If so, update the particle position; if not, proceed to the next judgment; 10) If the evolution is not in the same direction, further judge the calculated iteration number trigger probability function trigger_i. If trigger_i is less than zero, it is not triggered and the particle position is directly updated; if trigger_i is greater than zero, the next judgment is made; 11) If trigger_i is greater than zero, further judge the calculated simulation time trigger probability function trigger_h. If trigger_h is less than zero, it is not triggered and the particle position is directly updated; If trigger_h is greater than zero, trigger the evolution direction correction function, reverse the velocity direction, and make the particle evolution direction towards the direction where the energy storage power evolution direction is the same as the currently pre-judged energy storage action direction in step 1, that is: V(t) = -V(t) (8) 12) After correcting the particle evolution direction, update the particle position, that is: X(i + 1) = X(i) + V(i + 1) (9) 13) After updating the particle position, perform the operation in step 3 again to calculate the fitness function; 14) Judge whether the current iteration number i is greater than the iteration number value set by the simulation parameter. If the iteration number is not reached, return to step 5); if the iteration number is reached, output the optimal solution of the energy storage configuration and the operation ends.

2. The energy storage configuration optimization method considering the energy storage SOC constraint according to claim 1, wherein In step 1: In the current photovoltaic and energy storage power station, the actual output of the photovoltaic needs to follow the planned output curve. When the actual output of the photovoltaic exceeds the planned output curve, the probability that the current energy storage charging behavior is the optimal solution is greater; when the actual output of the photovoltaic is lower than the planned output curve, the probability that the current energy storage discharging behavior is the optimal solution is greater, which is expressed as: Wherein, PV is the actual output of the photovoltaic; PV f is the planned output of the photovoltaic; E op-fc = -1 represents that the probability that the current energy storage action is charging as the optimal solution is greater; E op-fc = 1 represents that the probability that the current energy storage action is discharging as the optimal solution is greater.

3. The energy storage configuration optimization method considering the energy storage SOC constraint according to claim 1, characterized in that In step 2, the set actual energy storage action direction is expressed as: Wherein, P(t) is the random particle of the particle swarm algorithm and also the energy storage operation power value; E op-re = 1 indicates that the energy storage operation is discharging; E op-re = -1 indicates that the energy storage operation is charging.

Images