Energy storage scheduling method and system of photovoltaic energy storage system based on particle swarm optimization

Abstract

The invention provides an energy storage scheduling method and system for a photovoltaic energy storage system based on particle swarm optimization, and the method comprises the steps: selecting a point with the maximum rise rate in an energy storage capacity-maximum annual average peak clipping proportion curve, and determining the initial capacity of the energy storage system; the charge state, the load power, the photovoltaic power generation power and the time of the energy storage system are used as state spaces, and the energy storage power is used as an optimization action; training a particle swarm optimization algorithm by taking the product minimization of the maximum grid power peak value square root in the time period and the daily grid power peak value square sum as an optimization target; in the range of + / -A% of the initial capacity value, energy storage capacity alternative values are selected at equal intervals, traversal is conducted through a trained particle swarm optimization algorithm, solving is conducted with the maximum weighted sum of the internal return rate IRR and the cost recovery rate as an objective function, and then an energy storage capacity configuration value and a corresponding action strategy are obtained. Accurate control over the energy storage capacity and the charging and discharging power is achieved.

Description

Technical Field

[0001] The present invention relates to the field of microgrid scheduling, and in particular to an energy storage scheduling method and system for a photovoltaic energy storage system based on particle swarm optimization. Background Art

[0002] Microgrids, as an emerging form of energy management, have experienced rapid growth in recent years. Providing a rational energy scheduling strategy is crucial for ensuring the safe, stable, and economical operation of microgrid systems. Since renewable energy exhibits natural complementarity in time and space, its rational use can effectively reduce energy storage capacity allocation, and energy storage devices can smooth out the random fluctuations caused by renewable energy access. The photovoltaic and energy storage management system uses regulatory measures to schedule the system to ensure stable operation, reduce operating costs, and achieve optimal energy management. Optimal scheduling in the photovoltaic and energy storage management system involves scheduling the photovoltaic and energy storage units in the microgrid to maximize the utilization of renewable energy, achieve peak load shifting, and achieve the goals of minimizing overall microgrid operating costs and maximizing overall social benefits.

[0003] Currently, most existing energy storage strategies rely on peak-valley arbitrage with fixed time and power. This approach maximizes economic benefits by storing electricity during periods of low electricity prices and discharging it during periods of high prices. However, this fixed time and power strategy ignores the volatility and flexibility of electricity demand and is unable to cope with real-time changes in electricity load. Summary of the Invention

[0004] The present invention proposes a method and system for energy storage scheduling of a photovoltaic energy storage system based on particle swarm optimization to solve the technical problem that existing energy storage scheduling schemes are difficult to cope with real-time changes in power load.

[0005] To solve the above technical problems, the present invention provides a method and system for energy storage scheduling of a photovoltaic energy storage system based on particle swarm optimization, comprising the following steps: Step S1: Select the point with the largest rising rate in the energy storage capacity-maximum annual average peak shaving ratio curve to determine the initial capacity of the energy storage system; Step S2: The energy storage system's state of charge, load power, photovoltaic power generation, and time are used as the state space, and the energy storage power is used as the optimization action. The particle swarm optimization algorithm is trained with the minimization of the product of the square root of the maximum grid power peak value within the time period and the sum of the squares of the daily grid power peak values ​​as the optimization objective. Step S3: Within the range of ±A% of the initial capacity value, select energy storage capacity alternative values ​​at equal intervals, traverse them using the trained particle swarm optimization algorithm, and solve the problem with the maximum weighted sum of the internal rate of return (IRR) and the cost recovery rate as the objective function, so as to obtain the energy storage capacity configuration value and the corresponding action strategy, where A is a number greater than zero.

[0006] Preferably, step S1 includes: Step S11: drawing a power curve; Step S12: taking the average value of the load power as the minimum value of the peak power, and taking the maximum value of the load power as the maximum value of the peak power; Step S13: Calculate the maximum peak shaving ratio under different energy storage capacities, plot the ideal peak power of each month under different energy storage capacities, and average the ideal peak shaving ratios of each month to obtain the maximum annual average peak shaving ratio; Step S14: According to the energy storage capacity-maximum annual average peak shaving ratio curve, the point with the largest rising rate is selected as the initial capacity of the energy storage system.

[0007] Preferably, the expression of the optimization objective in step S2 is: ; Where, represents the optimization objective; is the time length of the typical load power curve; Indicates the daily peak power input to a home.

[0008] Preferably, in step S2, state of charge constraints and charge and discharge power limits are imposed during training; the state of charge constraints are limited by the initial capacity.

[0009] Preferably, the state of charge constraint condition includes the maximum SOC value allowed during charging and the minimum SOC value allowed during discharging; the charge and discharge power limit includes the maximum charging power P CHG and maximum discharge power P DCHG .

[0010] Preferably, in step S2, when the particle swarm optimization algorithm is trained, the update expressions of position and velocity are: ; Where, i Number the particles; for speed; for location; represents the inertia weight; is the self-learning factor; is the group learning factor; For the i The optimal position of particles; is the historical optimal position of all particles; and A random number between 0 and 1.

[0011] Preferably, the objective function in step S4 is The expression is: ; Where, and is the weight coefficient; represents the internal rate of return; Represents the cost recovery rate.

[0012] Preferably, the calculation expression of the internal rate of return IRR is: ; ; ; ; ; ; Where, represents the initial investment; represents the unit capacity cost of energy storage; Indicates energy storage capacity; Indicates the battery operating life; Indicates the maximum cycle capacity; Indicates the i Months of cycle power; M Indicates the corresponding i Number of days in a month; and Indicates the corresponding i The first of the month j The depth and number of charge and discharge cycles per day; Indicates the project cycle; Indicates the cost of each battery replacement; Indicates annual income; For the i Maximum load peak reduction ratio of the month; For the i Months of maximum load; is the demand price; represents the maintenance cost; Indicates the battery replacement cost.

[0013] Preferably, the cost recovery rate is expressed as: ; ; Where, Represents the total cost.

[0014] The present invention also provides an energy storage scheduling system for a photovoltaic energy storage system based on particle swarm optimization, comprising: one or more processors and memories, and one or more programs, wherein the one or more programs are stored in the memories and are configured to be executed by the one or more processors, and the one or more programs include methods for executing the above-mentioned method.

[0015] The beneficial effects of the present invention include at least the following: first, based on historical data, the present invention analyzes the maximum peak-shaving ratios corresponding to different energy storage capacities to obtain an initial value of the energy storage capacity; then, using the existing power curve as a training sample set, the particle swarm algorithm is applied to comprehensively consider the battery's state of charge (SOC), load power, and relative time at the current moment to construct a basis function and determine the optimization target, thereby obtaining real-time control instructions for the energy storage system's charge and discharge power; and then, based on the instructions, the energy storage capacity is optimized, effectively solving the dynamic programming problem of charge and discharge control of the battery energy storage system, providing an innovative solution for demand management in the microgrid field, and achieving precise control of the energy storage capacity and charge and discharge power. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 Schematic diagram of a method flow in an embodiment of the present invention; Figure 2 This is a schematic diagram showing the principle of connecting a photovoltaic energy storage system to a campus microgrid according to an embodiment of the present invention; Figure 3 Schematic diagram of the achievable peak power value range of an embodiment of the present invention; Figure 4 This is a schematic diagram of a power curve of a certain park in January according to an embodiment of the present invention; Figure 5 The ideal value for January in the embodiment of the present invention is Schematic diagram of the curve; Figure 6 Schematic diagram of maximum demand in each month corresponding to different energy storage capacities according to an embodiment of the present invention; Figure 7 Schematic diagram of the annual average peak shaving ratio of different energy storage capacities according to an embodiment of the present invention; Figure 8 This is a schematic diagram of the maximum demand results for each month calculated according to an embodiment of the present invention. DETAILED DESCRIPTION

[0017] The following is a clear and complete description of the technical solutions in the embodiments of the present invention, in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts are within the scope of protection of the present invention.

[0018] like Figure 1 As shown, an embodiment of the present invention provides an energy storage scheduling method for a photovoltaic energy storage system based on particle swarm optimization, comprising the following steps: Step S1: Select the point with the largest rising rate in the energy storage capacity-maximum annual average peak shaving ratio curve to determine the initial capacity of the energy storage system.

[0019] Specifically, the principle of connecting the photovoltaic energy storage system to the park microgrid is as follows: Figure 2 shown.

[0020] Consider the comparative analysis of different energy storage capacities for demand management and optimize the energy storage capacity. The charging and discharging action of energy storage is reflected in the power curve as peak shaving and valley filling, which is reflected in the equivalent user load power curve. It can be seen that the achievable peak power The minimum value is the average load power , the achievable peak power The maximum value is the maximum load power ,like Figure 3 shown.

[0021] curve At peak power The integral of the part above the value is the The minimum energy storage capacity corresponding to the value , that is, we can get the ideal Parameter combination. In the embodiment of the present invention, the daily power curve of a certain park in January is as follows Figure 4 As shown, its Parameter curves such as Figure 5 Summarize the calculation results of different energy storage capacities for each month, as shown in Figure 6 shown.

[0022] In this embodiment, when calculating the peak shaving ratio, if the energy storage capacity is known, , according to the ideal After curve interpolation, the ideal peak power corresponding to this capacity is obtained, and then the maximum peak cutting ratio is obtained. : ; Plot the ideal peak power of each month under different energy storage capacities, and average the ideal peak shaving ratios of each month to obtain the maximum annual average peak shaving ratio, such as Figure 7 shown.

[0023] According to the energy storage capacity-maximum annual average peak shaving ratio curve, select the point with the largest rising rate, which is the optimal value of the energy storage capacity, and determine the initial capacity. .

[0024] Step S2: The state of charge of the energy storage system, load power, photovoltaic power generation power, and time are used as the state space, and the energy storage power is used as the optimization action. The particle swarm optimization algorithm is trained with the minimization of the product of the square root of the maximum grid power peak value in the time period and the sum of the squares of the daily grid power peak values ​​as the optimization objective.

[0025] Specifically, according to the energy storage system design parameters, namely the initial capacity, the SOC constraint conditions are clarified, namely the maximum SOC allowed during charging and the minimum SOC allowed during discharging; according to the PCS capacity and battery parameters of the energy storage system, the charge and discharge power limits, namely the maximum charging power P are clarified. CHG , maximum discharge power P DCHG .

[0026] Define the state space: ,in for t SOC at all times, for t Load power at the moment (per unit), for t Photovoltaic power generation at the moment (per unit value), for t The time of day at the moment.

[0027] Define action: store energy As an action to be optimized, the action should be related to the state: ; in For the status x Related basis functions, W is the weight vector. There are many choices for basis functions. To facilitate engineering implementation, we choose a polynomial function. Select the first-order polynomial form: . W It is a 5×1 vector, which is also the parameter that the particle swarm optimization PSO algorithm needs to optimize.

[0028] According to experience, the weight vector W Initialize, select W The typical value of each element of is [-10, 10], and the typical value of the search step range is [-1, 1].

[0029] Afterwards, typical photovoltaic power generation and power load power data of a certain period of time are selected as the training sample set. The duration of the training sample set is In order to ensure that the training sample set has a certain coverage capability, the total time length of the sample set in this example is The data interval is 5 minutes and the optimization index is defined as the minimization of the product of the square root of the maximum grid power peak value and the sum of the squares of the daily grid power peak values ​​during this period, that is: ; Where, represents the optimization objective; is the time length of the typical load power curve. In this embodiment Take 365; Indicates the daily peak power input to a home.

[0030] According to the PSO algorithm, a loop iteration is performed based on the training sample set. The update formula of its position (i.e. the parameter to be optimized) and speed (i.e. the search step length) is: ; Where, i Number the particles; for speed; for location; represents the inertia weight; is the self-learning factor; is the group learning factor; For the i The optimal position of particles; is the historical optimal position of all particles; and A random number between 0 and 1.

[0031] When the number of iterations reaches the set number, the algorithm is considered to have converged, and the real-time power control algorithm for the energy storage system is obtained.

[0032] Step S3: Within the range of ±A% of the initial capacity value, select alternative energy storage capacity values ​​at equal intervals, traverse them using the trained particle swarm optimization algorithm, and solve the problem with the maximum weighted sum of the internal rate of return (IRR) and the cost recovery rate as the objective function, thus obtaining the energy storage capacity configuration value and the corresponding action strategy, where A is a number greater than 0.

[0033] Specifically, in the embodiment of the present invention, an energy storage economic model is constructed, which mainly includes an energy storage system cost model, a revenue model, and an internal rate of return (IRR) model. The cost model calculation formula is: ; in is the unit capacity cost of energy storage, in yuan / kWh, Indicates energy storage capacity.

[0034] In order to better express the number of battery replacements and life, the rain flow counting method is used to build a battery life model. The model is to obtain the maximum cycle capacity based on the nominal parameters of the battery. , using the rain flow counting method, the depth of all charge and discharge cycles in each month can be calculated and the corresponding number of times , and then we can get the monthly cycle power : ; Where M represents the number of days corresponding to the i-th month; and Indicates the corresponding i The first of the month j The depth and number of charge and discharge cycles per day.

[0035] The annual cycle power is obtained by accumulating the monthly discharge power, and then the battery life is obtained: ; During the project life cycle Considering the replacement of battery cells, the battery replacement cost is: ; in The cost of each battery replacement (yuan).

[0036] The total cost model is: ; Where, Indicates maintenance cost.

[0037] The revenue model is the equivalent revenue after the demand charge is reduced, and the calculation formula is: ; Indicates annual income; For the i Maximum load peak reduction ratio of the month; For the i Months of maximum load; The demand price.

[0038] The annual cost recovery rate model is recorded as : ; The IRR calculation is based on discounting cash flows to make the net present value (NPV) equal to zero, where the cash flows include initial investment, operation and maintenance costs, and revenue. The calculation formula is as follows: ; in The initial investment takes a negative value; The annual net income of the project, including annual revenue and annual operation and maintenance costs; The cost of replacing the battery during its life cycle.

[0039] The objective function is: ; Where, and is the weight coefficient.

[0040] After that, the initial value of energy storage capacity Within the range of ±A, select the energy storage capacity alternative value at equal intervals, repeat the iteration, traverse the operation, and calculate the objective function. J By selecting the maximum value among the values, the optimized energy storage capacity configuration value and its action strategy can be obtained. In this embodiment, the value of A is 50%.

[0041] For a typical industrial park, the installed photovoltaic capacity is 4.5MWp, and the optimal energy storage capacity is 1MWh. Figure 8 The figure shows the maximum monthly demand after applying this intelligent peak-shaving algorithm. Under ideal conditions, the maximum peak-shaving ratio for a PV+1MWh energy storage system is 14.21%. The annual average peak-shaving ratio for a PV+storage system based on the PSO peak-shaving algorithm is 9.23%, and the annual average peak-shaving ratio for a fixed-value peak-shaving system is 4.97%.

[0042] The technical features of the above embodiments may be combined in any manner. To simplify the description, not all possible combinations of the technical features in the above embodiments are described. Only preferred embodiments of the present invention are presented. While the description is relatively specific and detailed, it should not be construed as limiting the scope of the present invention. As long as there are no conflicts in the combination of these technical features, they should be considered to be within the scope of this specification.

[0043] It should be noted that, for those skilled in the art, various modifications and improvements can be made without departing from the scope of the present invention, and these modifications and improvements fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be determined by the appended claims.

Claims

1. A method for energy storage scheduling of a photovoltaic energy storage system based on particle swarm optimization, characterized by: The following steps are involved: Step S1: Select the point with the largest rising rate in the energy storage capacity-maximum annual average peak shaving ratio curve to determine the initial capacity of the energy storage system; Step S2: The energy storage system's state of charge, load power, photovoltaic power generation, and time are used as the state space, and the energy storage power is used as the optimization action. The particle swarm optimization algorithm is trained with the minimization of the product of the square root of the maximum grid power peak value within the time period and the sum of the squares of the daily grid power peak values ​​as the optimization objective. Step S3: Within the range of ±A% of the initial capacity value, select energy storage capacity alternative values ​​at equal intervals, traverse them using the trained particle swarm optimization algorithm, and solve the problem with the maximum weighted sum of the internal rate of return (IRR) and the cost recovery rate as the objective function, so as to obtain the energy storage capacity configuration value and the corresponding action strategy, where A is a number greater than zero.

2. The energy storage scheduling method for a photovoltaic energy storage system based on particle swarm optimization according to claim 1, characterized in that: Step S1 includes: Step S11: drawing a power curve; Step S12: taking the average value of the load power as the minimum value of the peak power, and taking the maximum value of the load power as the maximum value of the peak power; Step S13: Calculate the maximum peak shaving ratio under different energy storage capacities, plot the ideal peak power of each month under different energy storage capacities, and average the ideal peak shaving ratios of each month to obtain the maximum annual average peak shaving ratio; Step S14: According to the energy storage capacity-maximum annual average peak shaving ratio curve, the point with the largest rising rate is selected as the initial capacity of the energy storage system.

3. The energy storage scheduling method for a photovoltaic energy storage system based on particle swarm optimization according to claim 1, characterized in that: The expression of the optimization objective in step S2 is: ； Where, represents the optimization objective; is the time length of the typical load power curve; Indicates the daily peak power input to a home.

4. The energy storage scheduling method for a photovoltaic energy storage system based on particle swarm optimization according to claim 1, characterized in that: In step S2, during training, state of charge constraints and charge and discharge power limits are imposed; the state of charge constraints are limited by the initial capacity.

5. The energy storage scheduling method for a photovoltaic energy storage system based on particle swarm optimization according to claim 4, characterized in that: The state of charge constraint condition includes the maximum SOC value allowed during charging and the minimum SOC value allowed during discharging; the charge and discharge power limit includes the maximum charging power P CHG and maximum discharge power P DCHG .

6. The energy storage scheduling method for a photovoltaic energy storage system based on particle swarm optimization according to claim 1, characterized in that: In step S2, when the particle swarm optimization algorithm is trained, the update expressions of position and velocity are: ； Where, i Number the particles; for speed; for location; represents the inertia weight; is the self-learning factor; is the group learning factor; For the i The optimal position of particles; is the historical optimal position of all particles; and A random number between 0 and 1.

7. The energy storage scheduling method for a photovoltaic energy storage system based on particle swarm optimization according to claim 1, characterized in that: The objective function in step S4 The expression is: ； Where, and is the weight coefficient; represents the internal rate of return; Represents the cost recovery rate.

8. The energy storage scheduling method for a photovoltaic energy storage system based on particle swarm optimization according to claim 7, characterized in that: The calculation expression of the internal rate of return IRR is: ； ； ； ； ； ； Where, represents the initial investment; represents the unit capacity cost of energy storage; Indicates energy storage capacity; Indicates the battery operating life; Indicates the maximum cycle capacity; Indicates the i Months of cycle power; M represents the number of days corresponding to the i-th month; and Indicates the corresponding i The first of the month j The depth and number of charge and discharge cycles per day; Indicates the project cycle; represents the cost of each battery replacement; Indicates annual income; For the i Maximum load peak reduction ratio of the month; For the i Maximum load for the month; is the demand price; represents the maintenance cost; Indicates the battery replacement cost.

9. The energy storage scheduling method for a photovoltaic energy storage system based on particle swarm optimization according to claim 8, characterized in that: The expression of the cost recovery rate is: ； ； Where, Represents the total cost.

10. An energy storage scheduling system for a photovoltaic energy storage system based on particle swarm optimization, characterized by: include: One or more processors and memory, and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include instructions for executing the method according to any one of claims 1 to 9.

Images