Method for optimizing hybrid photovoltaic energy storage based on PerturbandObserve algorithm

Abstract

The invention relates to the technical field of photovoltaic energy storage methods, and discloses a method for optimizing hybrid photovoltaic energy storage based on a Perturbobserve algorithm, which comprises a load, a photovoltaic power generation subsystem, an energy storage subsystem, a control and management system and a grid-connected system. The specific control method comprises the following steps: S1, obtaining an illumination intensity value within one day through a meteorological station, and obtaining a power grid net load value within one day through a voltage sensor and a current sensor; s2, taking economy optimization as a target, considering photovoltaic energy storage capacity constraints, and establishing an energy storage SOC control strategy target function; and S3, solving is carried out by using a PerturBobserve algorithm, and an optimal control pulse sequence is obtained. By optimizing the charging and discharging time and intensity of the energy storage battery, the consumption rate of photovoltaic power generation is maximized, the light abandoning phenomenon is reduced, and the energy loss in the charging and discharging process is reduced. Meanwhile, the PertuBobserve algorithm can avoid the excessive charging and discharging of the energy storage battery while considering the system load, prolongs the service life of the battery, and solves the problem of insufficient management of the service life of the battery in a traditional method.

Description

Technical Field

[0001] The present invention relates to the field of photovoltaic energy storage methods, and in particular to a hybrid photovoltaic energy storage method based on PerturbandObserve algorithm optimization. Background Art

[0002] With the advancement of renewable energy technology, photovoltaic power generation, as a representative of clean energy, is driving the transformation of the global energy structure. However, photovoltaic power generation is affected by light conditions and is volatile and intermittent, resulting in a mismatch between energy supply and demand when connected to the grid, affecting the stability of power supply. In order to solve this problem, photovoltaic energy storage systems have emerged. Energy storage batteries store electricity when there is sufficient light and release electricity when there is insufficient light, thereby balancing supply and demand and improving the stability and efficiency of the system.

[0003] The control strategies of traditional photovoltaic energy storage systems mostly use simple scheduling methods based on rules or experience, such as preset charging and discharging schedules or maximum power point tracking (MPPT) technology. Although these methods can achieve basic operation control, they have limited optimization space and are difficult to adapt to complex and changing operating environments, such as high-load periods or continuous cloudy days. In addition, traditional methods lack sufficient consideration of the life cycle of energy storage batteries, which can easily lead to overcharging or over-discharging of batteries, affecting battery efficiency and service life.

[0004] In order to solve the above problems, this application proposes a hybrid photovoltaic energy storage method based on the PerturbandObserve algorithm. Summary of the invention

[0005] 1. Purpose of the invention In order to solve the technical problems existing in the background technology, the present invention proposes a hybrid photovoltaic energy storage method based on the PerturbandObserve algorithm. The present invention maximizes the absorption rate of photovoltaic power generation, reduces the phenomenon of abandoned light, and reduces the energy loss during the charging and discharging process by optimizing the charging and discharging timing and intensity of the energy storage battery. At the same time, while taking the system load into consideration, the PerturBObserve algorithm can also avoid excessive charging and discharging of the energy storage battery, extend the battery life, and solve the problem of insufficient battery life management in traditional methods. Through these optimization effects, the operating efficiency of the system has been significantly improved, and both the economy and the clean energy absorption rate have been improved.

[0006] (II) Technical solution To solve the above problems, the present invention provides a hybrid photovoltaic energy storage method based on the PerturbandObserve algorithm, including a load, a photovoltaic power generation subsystem, an energy storage subsystem, a control and management system, and a grid-connected system; The specific control methods are as follows: S1. Obtain the light intensity value within a day through the weather station, and obtain the net load value of the power grid within a day through the voltage and current sensors; S2. Taking economic optimization as the goal, considering the photovoltaic energy storage capacity constraints, establish the energy storage SOC control strategy objective function; S3. Use the PerturBObserve algorithm to solve and obtain the optimal control pulse sequence.

[0007] Preferably, in S1: The key parameters of the system are collected through the weather station and voltage and current sensors to provide a data basis for subsequent optimization control. Figure 3 As shown; include: S1.1. Obtain the light intensity value within a day through the weather station, and determine the characteristic relationship of the photovoltaic system by combining the current and voltage output by the photovoltaic array. The output current PV of the photovoltaic array is related to parameters such as light intensity S(t), photovoltaic array temperature, and internal circuit impedance, and can be expressed by the following formula: ; Among them, PV is the output current of the photovoltaic array, U(t) is the output voltage of the photovoltaic array, qPD is the photovoltaic conversion current coefficient, KPD is the photovoltaic output current variation coefficient related to temperature, T is the current temperature of the photovoltaic array, T0 is the reference temperature, Rd and Rc are the series shunt resistance and parallel shunt resistance respectively, ε is the Boltzmann constant, α is the ratio of the shunt resistance to the equivalent internal resistance of the battery, and S(t) is the light intensity. The goal of the formula is to derive the real-time output current of the photovoltaic system through known environmental parameters and circuit characteristics, and provide input basis for the scheduling of the energy storage system. For example, when the light intensity S(t) increases, PV will rise significantly, and the increase in the series shunt resistance will cause PV to drop.

[0008] S1.2. Use voltage and current sensors to obtain the net load change value W(t) of the power grid. The net load of the power grid reflects the dynamic changes in the load demand on the system, and its value is affected by the grid voltage, the SOC of the energy storage system, and the internal loss of the system. The net load can be expressed as: ; Here, W(t) represents the net load of the grid at the current moment t, W(t-1) is the net load at the previous moment, U(t) and U(t-1) are the grid-connected voltages at the current and previous moments, respectively, IG(t) and IG(t-1) are the energy storage system SOCs at the current and previous moments, respectively, G is the total system capacity, and ηg is the grid loss coefficient. Through this formula, the net load state of the grid at different moments can be dynamically captured, and the impact of system losses on the overall performance can be taken into account. The effect of the formula is to accurately describe the dynamic grid state, thereby providing data support for the subsequent optimization of the energy storage SOC control strategy. For example, when IG(t) is significantly greater than IG(t-1), it means that the energy storage system is in a discharge state and the net load is reduced.

[0009] Through the above two formulas, the system can describe the output characteristics of the photovoltaic array and the dynamic changes of the net load of the power grid. The acquisition of these key parameters is directly related to the solution of the subsequent optimization objective function and the design of the control pulse sequence. For example, during the noon period on a sunny day, S(t) reaches its maximum value and PV will rise sharply. By adjusting IG(t), the dynamic demand of the power grid can be adapted, thus achieving a balance between economy and safety.

[0010] Preferably, S2 includes: The core of S2 is to establish and solve the optimal control model based on the actual operating characteristics of the photovoltaic energy storage system coupled with the power grid, comprehensively consider the loss and economy, and thus realize the optimal charging and discharging strategy design of the energy storage system. Figure 4 As shown, first, in S2.1, an optimization model is established based on the characteristics of the photovoltaic energy storage system. The model takes economic optimization as the goal and uses the optimization of control variables to balance the energy interaction between photovoltaic energy storage and the power grid.

[0011] S2.2, use the PerturBObserve optimization algorithm to optimize the model, and the objective function is the minimum value of the charging and discharging cost at each moment. The economic objective function is expressed as: ; Among them, Fmax is the charging and discharging cost required by the energy storage system when the SOC reaches the maximum value, Fdischarging is the cost of the discharge stage, and Fcharging is the cost of the charging stage. The calculation of the charging and discharging cost needs to take into account the efficiency factor, among which the charging efficiency ηc and the discharging efficiency ηd are affected by the internal parameters of the system. The purpose of the objective function is to reduce the overall operating cost by optimizing the charging and discharging process. For example, when the photovoltaic power generation is large in a certain period of time, by increasing the utilization of the charging efficiency ηc, the cost of purchasing electricity from the power grid can be effectively reduced, thereby optimizing the economy.

[0012] In S2.3, the SOC constraint model is introduced to limit the SOC range during the optimization process to ensure the safety and stability of system operation. The expression of SOC constraint is: ; Among them, Gm is the actual charge and discharge amount of the system in one day, G is the total capacity of the system, IGmax and IGmin are the upper and lower limits of SOC respectively. Through this constraint, it can be ensured that SOC is always within the safe range during the entire optimization process to avoid overcharging or over-discharging. For example, when photovoltaic power cannot generate electricity at night, the lower limit of SOC IGmin can avoid deep discharge of the energy storage system, while the upper limit of SOC IGmax can limit overcharging during the day.

[0013] In summary, the global optimal control of the energy storage system can be achieved by combining the PerturBObserve optimization algorithm with the economic objective function and the SOC constraint model. For example, on a certain day, assuming that the total system capacity is 100 kWh, the upper and lower limits of SOC are 10% and 90%, respectively, and the charging and discharging efficiencies are 0.95 and 0.9, respectively. When there is excess photovoltaic power generation during the day, the SOC is optimized to 80% to reduce discharge losses, and high-efficiency discharge is used to maintain stable operation of the power grid during peak loads at night, thereby achieving the dual goals of economy and safety.

[0014] Preferably, in S3: The core of S3 is to solve the optimal control pulse sequence of the energy storage system through the PerturBObserve algorithm to achieve economic and operational efficiency improvement. As shown in Figure 5. The process includes initialization, disturbance sequence generation, control sequence correction and the output of the final optimization result, as follows: S3.1. Initialize the parameters required by the algorithm. First, randomly determine the decision variables and clearly optimize the objective function, which is closely related to the economic objective function and SOC constraint model established in S2. At the same time, set the values ​​of the disturbance pulse width t and sequence length m in the control strategy, and give the initial state variables x0, such as the initial SOC value and the current photovoltaic power output.

[0015] S3.2. Set the algorithm's precision index εp and error index εv, and define the random perturbation pulse intensity function f(r), where r is a random variable in the range [0,1]. This intensity function determines the randomness of the perturbation sequence and affects the algorithm's convergence speed and result accuracy. For example, f(r) can be set to a uniform distribution or a normal distribution to adapt to different system dynamic characteristics.

[0016] S3. Calculate the number of samples and generate a sample set S(x), x∈x0, to form the algorithm input sample space. The number of samples determines the solution range of the algorithm and the degree of approximation to the complex objective function.

[0017] S3.5 - S3.7. Generate the perturbation sequence and the control sequence through the algorithm loop variable. The calculation formula for the perturbation pulse intensity xi is as follows: ; where b is the offset value within the random variable range [-1, 1], is the inverse function of the perturbation intensity. The goal of this formula is to generate a random perturbation sequence to explore the neighborhood space of the optimal solution through the perturbation characteristics. On this basis, the control sequence is obtained through weighted calculation: ; Here, w is the control weight coefficient, α is the sequence mean, and K is the length of the optimized sequence. Through the weighted generation of the control sequence, the optimal solution can be gradually approximated, reducing system losses and improving stability.

[0018] S3.8 - S3.9. Evaluate the current control effect by calculating the mean and covariance of the perturbation sequence and the control sequence. The deviation calculation formula is: ; where is the mean of the perturbation sequence and is the mean of the control sequence. The calculation of covariance is used to judge the correlation between the perturbation sequence and the control sequence, guiding the optimization direction; When the algorithm loop variable n < Ns, return to S3.6 to continue generating a new perturbation sequence and correcting the control sequence; when the sampling number reaches Ns, judge whether the optimization condition is satisfied according to the accuracy index in S3.1. If the condition is satisfied, output the final control sequence; otherwise, return to S3.3 to readjust the parameters.

[0019] Preferably, it further includes: Calculate the operating cost of the hybrid photovoltaic energy storage system through the optimized control pulse sequence. For example, assume that the sampling number within a day is 200, the length m of the perturbation sequence is 10, the perturbation pulse width t is 20, and the system realizes an average charging efficiency improvement of 5% and an economic cost reduction of 15% through optimization. The final control pulse sequence balances economy and system safety in the charge and discharge process by correcting the deviation and covariance.

[0020] In the process of optimizing the photovoltaic energy storage system based on the PerturBObserve algorithm, adding the energy storage SOC constraint is a key link to achieve the balance between system safety and economy. SOC (State of Charge of the energy storage battery) reflects the current charge and discharge state of the energy storage battery and is an important indicator to measure the available performance of the energy storage device. The SOC constraint can ensure that the energy storage system avoids overcharging and over-discharging during operation, while improving the charge and discharge efficiency and reducing the operating cost; In order to incorporate SOC into the optimization model, it is necessary to introduce a dynamic expression of SOC into the control strategy and define the upper limit of SOC. and lower limit As a constraint, the effect of the energy storage time interval Δt on the SOC change is also considered. The dynamic change of SOC can be described by the following formula: ; in, is the energy storage state at time t+1, is the energy storage state at time t, and are charging power and discharging power respectively, ηc and ηd are charging efficiency and discharging efficiency respectively, is the total capacity of the battery. The goal of this formula is to calculate the change of SOC in real time, combine power input and system efficiency, and dynamically adjust the operating state of the energy storage system.

[0021] In the optimization control model, the objective function with SOC as constraint can be expressed as: ; Among them, F is the total charging and discharging cost in one day, and are the charging and discharging costs in the tth period, respectively. The constraints ensure that the SOC operates within a safe range. By introducing SOC constraints, the system can be prevented from over-discharging during high-load periods or over-charging during low-load periods, thereby extending the life of the energy storage battery and improving the system economy.

[0022] Taking practical application as an example, assuming that the total battery capacity of a photovoltaic energy storage system is 100 kWh, the charging efficiency is 95%, and the discharging efficiency is 90%. and Set to 20% and 90% respectively. In one day of operation, the peak power of photovoltaic power generation is 50 kW and the peak load demand is 40 kW. When there is excess photovoltaic power generation at noon, the SOC gradually approaches 90%, and the system automatically reduces the charging power to prevent overcharging; and when the load demand increases at night, the SOC gradually decreases to 25%, and the discharge power is limited by Constraints to avoid deep discharge that may damage the battery.

[0023] By introducing the above formulas and constraints, combined with the PerturBObserve algorithm, the grid-connected control strategy of the photovoltaic energy storage system can be effectively optimized, ensuring system safety and long-term use of equipment while meeting operational economy.

[0024] Preferably, in S2, the input of the optimization model includes the number of time series sampling points T, the variable μt of the optimization control sequence, the mean of the control item, the length of the disturbance sequence m, the length of the optimization sequence K, and the correction coefficient σ. The goal of the optimization model is to minimize the operating loss of the energy storage system within a day and achieve the maximum utilization of photovoltaic power generation by adjusting the optimization control variable μt.

[0025] The above technical solution of the present invention has the following beneficial technical effects: Improve economy and reduce operating costs By minimizing the cost of the charging and discharging process as the optimization goal, combined with the SOC state constraints of the energy storage battery, the PerturBObserve algorithm can intelligently and dynamically adjust the storage and release timing of photovoltaic power, avoid unnecessary grid purchases and battery overcharging or over-discharging, and ensure that the energy storage system operates under the most economical conditions. Especially during peak grid load periods, the system can intelligently adjust the charging and discharging of energy storage batteries according to real-time power generation and load requirements, reduce dependence on traditional energy, and effectively reduce overall operating costs and energy expenditures.

[0026] Optimizing the life cycle of energy storage batteries The service life of energy storage batteries is directly affected by the number and depth of their charge and discharge. Overcharging or over-discharging will significantly reduce the effective life of the battery. Through the precise control of SOC by the PerturBObserve algorithm, the system can ensure that the energy storage battery always works in a reasonable charging and discharging range, avoid excessive use, and delay the decline of battery capacity. Long-term optimization management not only reduces the frequency of battery replacement, but also reduces battery-related maintenance costs and environmental pressures.

[0027] Improve photovoltaic power generation utilization and system stability Photovoltaic power generation is highly volatile, especially on cloudy days, at night, or during high-load periods, when insufficient power generation may occur. Through the algorithm's disturbance feedback mechanism, the system can adjust the charging and discharging behavior of the energy storage battery in real time, ensuring that even during the trough of photovoltaic power generation, the system can still provide stable power output and avoid excessive dependence on the power grid. In addition, the PerturBObserve algorithm can make optimal adjustments under different operating conditions, enhancing the system's ability to cope with sudden changes and fluctuating loads, and ensuring the stable operation of the system in different environments.

[0028] Improve the consumption rate and environmental benefits of clean energy By using this optimization control strategy, the system can maximize the use of surplus power from photovoltaic power generation, reduce the grid's dependence on traditional fossil energy, and increase the clean energy consumption rate. While meeting the grid's needs, it reduces the phenomenon of photovoltaic power generation abandonment, helps achieve a higher level of green energy utilization, reduces carbon emissions, and meets the goals of sustainable development and environmental protection. The system provides effective support for promoting the development of low-carbon energy and has significant social value.

[0029] Through these optimization effects, the hybrid photovoltaic energy storage system based on the PerturBObserve algorithm not only improves economy and battery life, but also significantly improves the utilization rate of clean energy and system stability, demonstrating high technological innovation and practical application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 This is a system diagram of a hybrid photovoltaic energy storage method optimized based on the PerturbandObserve algorithm proposed in the present invention.

[0031] Figure 2 This is a flow chart of a hybrid photovoltaic energy storage method based on the PerturbandObserve algorithm proposed in the present invention.

[0032] Figure 3 This is a data flow chart of a hybrid photovoltaic energy storage method optimized based on the PerturbandObserve algorithm proposed in the present invention.

[0033] Figure 4 This is a flowchart of step 2 in a method for optimizing hybrid photovoltaic energy storage based on the PerturbandObserve algorithm proposed in the present invention.

[0034] Figure 5 This is a system diagram of a hybrid photovoltaic energy storage method optimized based on the PerturbandObserve algorithm proposed in the present invention. DETAILED DESCRIPTION

[0035] In order to make the purpose, technical scheme and advantages of the present invention clearer, the present invention is further described in detail below in conjunction with specific embodiments and with reference to the accompanying drawings. It should be understood that these descriptions are only exemplary and are not intended to limit the scope of the present invention. In addition, in the following description, the description of well-known structures and technologies is omitted to avoid unnecessary confusion of the concept of the present invention.

[0036] like Figure 1-5 As shown, the present invention proposes a hybrid photovoltaic energy storage method based on the PerturbandObserve algorithm, including a load, a photovoltaic power generation subsystem, an energy storage subsystem, a control and management system, and a grid-connected system; Figure 1 As shown, it includes four major modules: photovoltaic power generation subsystem, energy storage subsystem, grid-connected subsystem, and control and management system. Each module is seamlessly connected through power transmission lines and communication networks. As the main energy supply module, the photovoltaic power generation subsystem converts light energy into direct current through photovoltaic modules, which is converted into alternating current through inverters and then input into the energy storage subsystem or grid-connected subsystem. The function of this module is to provide clean energy to reduce the use of traditional energy, while providing a stable basic energy input for subsequent modules; the energy storage subsystem stores excess energy of the photovoltaic power generation subsystem through efficient battery packs, or supplements power to the load and grid-connected subsystem when photovoltaic power generation is insufficient. This module dynamically adjusts the charging and discharging process to reduce peaks and fill valleys, smooth out energy fluctuations, and protects the battery by adding SOC (remaining state of charge of the energy storage battery) constraints; the grid-connected subsystem is connected to the power grid. The core module of the photovoltaic energy storage system and the entire photovoltaic energy storage system is responsible for realizing the two-way flow of electricity. It transmits electricity to external loads or draws insufficient energy from the grid through precise power control and grid synchronization technology. Its role is to ensure the coordinated operation of the system and the grid. The control and management system is the core module for achieving overall optimization. It collects key data such as the output power of the photovoltaic power generation subsystem, the SOC status of the energy storage subsystem, and the grid parameters of the grid-connected subsystem in real time, and uses the PerturBObserve algorithm to dynamically optimize the control strategy to generate the optimal charging and discharging instructions and grid-connected scheduling plan. Its main role is to coordinate the operation of each module to achieve global optimization of economy and safety.

[0037] The coordination between modules is achieved through real-time information exchange based on the communication network and an algorithm-driven decision-making system. The photovoltaic power generation subsystem monitors the light intensity and component operating status in real time through the control and management system, and transmits the power generation data to the energy storage subsystem and the grid-connected subsystem; the energy storage subsystem dynamically adjusts the charging and discharging power according to the SOC optimization target generated by the control and management system to ensure that the battery operates in a safe and efficient range; the grid-connected subsystem flexibly adjusts the output power according to the grid status and load demand under the unified dispatch of the control and management system to ensure the stability of power transmission. Through the collaborative work between the above modules, the system can solve the problems of insufficient economy, overcharge and over-discharge of energy storage batteries, and low energy utilization in traditional photovoltaic energy storage systems, and realize the optimization of energy dispatch and efficient and stable system operation.

[0038] The specific control methods are as follows: S1. Obtain the light intensity value within a day through the weather station, and obtain the net load value of the power grid within a day through the voltage and current sensors; collect the key parameters of the system through the weather station and voltage and current sensors to provide a data basis for subsequent optimization control. Figure 3 As shown; Specifically include: S1.1. Obtain the light intensity value within a day through the weather station, and determine the characteristic relationship of the photovoltaic system by combining the current and voltage output by the photovoltaic array. The output current PV of the photovoltaic array is related to parameters such as light intensity S(t), photovoltaic array temperature, and internal circuit impedance, and can be expressed by the following formula: Among them, PV is the output current of the photovoltaic array, U(t) is the output voltage of the photovoltaic array, qPD is the photovoltaic conversion current coefficient, KPD is the photovoltaic output current variation coefficient related to temperature, T is the current temperature of the photovoltaic array, T0 is the reference temperature, Rd and Rc are the series shunt resistance and parallel shunt resistance respectively, ε is the Boltzmann constant, α is the ratio of the shunt resistance to the equivalent internal resistance of the battery, and S(t) is the light intensity. The goal of the formula is to derive the real-time output current of the photovoltaic system through known environmental parameters and circuit characteristics, and provide input basis for the scheduling of the energy storage system. For example, when the light intensity S(t) increases, PV will rise significantly, and the increase in the series shunt resistance will cause PV to drop.

[0039] S1.2. Use voltage and current sensors to obtain the net load change value W(t) of the power grid. The net load of the power grid reflects the dynamic changes in the load demand on the system, and its value is affected by the grid voltage, the SOC of the energy storage system, and the internal loss of the system. The net load can be expressed as: Here, W(t) represents the net load of the grid at the current moment t, W(t-1) is the net load at the previous moment, U(t) and U(t-1) are the grid-connected voltages at the current and previous moments, respectively, IG(t) and IG(t-1) are the energy storage system SOCs at the current and previous moments, respectively, G is the total system capacity, and ηg is the grid loss coefficient. Through this formula, the net load state of the grid at different moments can be dynamically captured, and the impact of system losses on the overall performance can be taken into account. The effect of the formula is to accurately describe the dynamic grid state, thereby providing data support for the subsequent optimization of the energy storage SOC control strategy. For example, when IG(t) is significantly greater than IG(t-1), it means that the energy storage system is in a discharge state and the net load is reduced.

[0040] Through the above two formulas, the system can describe the output characteristics of the photovoltaic array and the dynamic changes of the net load of the power grid. The acquisition of these key parameters is directly related to the solution of the subsequent optimization objective function and the design of the control pulse sequence. For example, during the noon period on a sunny day, S(t) reaches its maximum value and PV will rise sharply. By adjusting IG(t), the dynamic demand of the power grid can be adapted, thus achieving a balance between economy and safety.

[0041] S2. Taking economic optimization as the goal, considering the photovoltaic energy storage capacity constraints, establish the energy storage SOC control strategy objective function; The core of S2 is to establish and solve the optimal control model based on the actual operating characteristics of the photovoltaic energy storage system coupled with the power grid, comprehensively consider the loss and economy, and thus realize the optimal charging and discharging strategy design of the energy storage system. Figure 4 shown.

[0042] S2.1. An optimization model is established based on the characteristics of the photovoltaic energy storage system. The model aims at economic optimization and uses the optimization of control variables to balance the energy interaction between photovoltaic energy storage and the power grid.

[0043] The input of the optimization model includes the number of time series sampling points T, the variable μt of the optimization control sequence, the mean of the control item, the length of the disturbance sequence m, the length of the optimization sequence K, and the correction coefficient σ. The goal of the optimization model is to minimize the operating loss of the energy storage system within a day and achieve the maximum utilization of photovoltaic power generation by adjusting the optimization control variable μt.

[0044] S2.2, use the PerturBObserve optimization algorithm to optimize the model, and the objective function is the minimum value of the charging and discharging cost at each moment. The economic objective function is expressed as: ; Among them, Fmax is the charging and discharging cost required by the energy storage system when the SOC reaches the maximum value, Fdischarging is the cost of the discharge stage, and Fcharging is the cost of the charging stage. The calculation of the charging and discharging cost needs to take into account the efficiency factor, among which the charging efficiency ηc and the discharging efficiency ηd are affected by the internal parameters of the system. The purpose of the objective function is to reduce the overall operating cost by optimizing the charging and discharging process. For example, when the photovoltaic power generation is large in a certain period of time, by increasing the utilization of the charging efficiency ηc, the cost of purchasing electricity from the power grid can be effectively reduced, thereby optimizing the economy.

[0045] S2.3, introduce the SOC constraint model to limit the SOC range in the optimization process and ensure the safety and stability of system operation. The expression of SOC constraint is: ; Among them, Gm is the actual charge and discharge amount of the system in one day, G is the total capacity of the system, IGmax and IGmin are the upper and lower limits of SOC respectively. Through this constraint, it can be ensured that SOC is always within the safe range during the entire optimization process to avoid overcharging or over-discharging. For example, when photovoltaic power cannot generate electricity at night, the lower limit of SOC IGmin can avoid deep discharge of the energy storage system, while the upper limit of SOC IGmax can limit overcharging during the day.

[0046] In summary, the global optimal control of the energy storage system can be achieved by combining the PerturBObserve optimization algorithm with the economic objective function and the SOC constraint model. For example, on a certain day, assuming that the total system capacity is 100 kWh, the upper and lower limits of SOC are 10% and 90%, respectively, and the charging and discharging efficiencies are 0.95 and 0.9, respectively. When there is excess photovoltaic power generation during the day, the SOC is optimized to 80% to reduce discharge losses, and high-efficiency discharge is used to maintain stable operation of the power grid during peak loads at night, thereby achieving the dual goals of economy and safety.

[0047] S3. Use the PerturBObserve algorithm to solve and obtain the optimal control pulse sequence.

[0048] The core of S3 is to solve the optimal control pulse sequence of the energy storage system through the PerturBObserve algorithm to achieve economic and operational efficiency improvement. As shown in Figure 5. The process includes initialization, disturbance sequence generation, control sequence correction and the output of the final optimization result, as follows: S3.1. Initialize the parameters required by the algorithm. First, randomly determine the decision variables and clearly optimize the objective function, which is closely related to the economic objective function and SOC constraint model established in S2. At the same time, set the values ​​of the disturbance pulse width t and sequence length m in the control strategy, and give the initial state variables x0, such as the initial SOC value and the current photovoltaic power output.

[0049] S3.2. Set the algorithm's precision index εp and error index εv, and define the random perturbation pulse intensity function f(r), where r is a random variable in the range [0,1]. This intensity function determines the randomness of the perturbation sequence and affects the algorithm's convergence speed and result accuracy. For example, f(r) can be set to a uniform distribution or a normal distribution to adapt to different system dynamic characteristics.

[0050] S3. Calculate the number of samples and generate a sample set S(x), x∈x0, to form the algorithm input sample space. The number of samples determines the solution range of the algorithm and the degree of approximation to the complex objective function.

[0051] S3.5-S3.7, generate disturbance sequence and control sequence through algorithm loop variables. The calculation formula of disturbance pulse intensity xi is as follows: ; Where b is the offset value of the random variable in the range [-1,1], is the inverse function of the perturbation intensity. The goal of this formula is to generate a random perturbation sequence and explore the neighborhood space of the optimal solution through the perturbation characteristics. On this basis, the control sequence is obtained through weighted calculation: ; Here, \(w\) is the control weight coefficient, \(\alpha\) is the sequence mean, and \(K\) is the optimized sequence length. By controlling the weighted generation of the sequence, the optimal solution can be gradually approximated, reducing system losses and improving stability.

[0052] S3.8 - S3.9, evaluate the current control effect by calculating the mean and covariance of the perturbation sequence and the control sequence. The deviation calculation formula is: ; where, \(\overline{x}\) is the mean of the perturbation sequence, and \(\overline{y}\) is the mean of the control sequence. The calculation of covariance is used to judge the correlation between the perturbation sequence and the control sequence, guiding the optimization direction; When the algorithm loop variable \(n < N_s\), return to S3.6, continue to generate a new perturbation sequence and correct the control sequence; when the sampling number \(N_s\) is reached, judge whether the optimization condition is satisfied according to the accuracy index in S3.1. If the condition is satisfied, output the final control sequence; otherwise, return to S3.3 to re - adjust the parameters.

[0053] Preferably, it further includes: Calculate the operating cost of the hybrid photovoltaic energy storage system through the optimized control pulse sequence. For example, assume that the sampling number within a day is 200, the length \(m\) of the perturbation sequence is 10, the width \(t\) of the perturbation pulse is 20, and the system realizes an average charging efficiency improvement of 5% and an economic cost reduction of 15% after optimization. The final control pulse sequence balances economy and system safety in the charge - discharge process by correcting the deviation and covariance.

[0054] In the process of optimizing the photovoltaic energy storage system based on the PerturBObserve algorithm, adding the energy storage SOC constraint is the key link to achieve the balance between system safety and economy. SOC (State of Charge of the energy storage battery) reflects the current charge - discharge state of the energy storage battery and is an important indicator to measure the available performance of the energy storage device. The SOC constraint can ensure that the energy storage system avoids over - charging and over - discharging during operation, while improving the charge - discharge efficiency and reducing the operating cost; To incorporate SOC into the optimization model, a dynamic expression of SOC needs to be introduced in the control strategy, defining the upper limit and the lower limit as constraint conditions, and at the same time considering the influence of the energy storage charging time interval \(\Delta t\) on the change of SOC. The dynamic change of SOC can be described by the following formula: ; where, \(S_{t + 1}\) is the energy storage state at time \(t + 1\), \(S_t\) is the energy storage state at time \(t\), and \(P_c\) and \(P_d\) are the charging power and discharging power respectively, \(\eta_c\) and \(\eta_d\) are the charging efficiency and discharging efficiency, is the total capacity of the battery. The goal of this formula is to calculate the change of SOC in real time, combine power input and system efficiency, and dynamically adjust the operating state of the energy storage system.

[0055] In the optimization control model, the objective function with SOC as constraint can be expressed as: ; Among them, F is the total charging and discharging cost in one day, and are the charging and discharging costs in the tth period, respectively. The constraints ensure that the SOC operates within a safe range. By introducing SOC constraints, the system can be prevented from over-discharging during high-load periods or over-charging during low-load periods, thereby extending the life of the energy storage battery and improving the system economy.

[0056] Taking practical application as an example, assuming that the total battery capacity of a photovoltaic energy storage system is 100 kWh, the charging efficiency is 95%, and the discharging efficiency is 90%. and Set to 20% and 90% respectively. In one day of operation, the peak power of photovoltaic power generation is 50 kW and the peak load demand is 40 kW. When there is excess photovoltaic power generation at noon, the SOC gradually approaches 90%, and the system automatically reduces the charging power to prevent overcharging; and when the load demand increases at night, the SOC gradually decreases to 25%, and the discharge power is limited by Constraints to avoid deep discharge that may damage the battery.

[0057] By introducing the above formulas and constraints, combined with the PerturBObserve algorithm, the grid-connected control strategy of the photovoltaic energy storage system can be effectively optimized, ensuring system safety and long-term use of equipment while meeting operational economy.

[0058] It should be understood that the above specific embodiments of the present invention are only used to illustrate or explain the principles of the present invention, and do not constitute a limitation of the present invention. Therefore, any modifications, equivalent substitutions, improvements, etc. made without departing from the spirit and scope of the present invention should be included in the protection scope of the present invention. In addition, the appended claims of the present invention are intended to cover all changes and modifications that fall within the scope and boundaries of the appended claims, or the equivalent forms of such scope and boundaries.

Claims

1. A hybrid photovoltaic energy storage method based on the PerturbandObserve algorithm, characterized in that: Including load, photovoltaic subsystem, energy storage subsystem, control and management system and grid connection system; Here’s how: S1. Obtain the light intensity value within a day through the weather station, and obtain the net load value of the power grid within a day through the voltage and current sensors; S2. Taking economic optimization as the goal, considering the photovoltaic energy storage capacity constraints, establish the energy storage SOC control strategy objective function; S3. Use the PerturBObserve algorithm to solve and obtain the optimal control pulse sequence.

2. The method for optimizing hybrid photovoltaic energy storage based on the PerturbandObserve algorithm according to claim 1, characterized in that: In S1, it includes: S1.

1. Obtain the light intensity value within a day through the meteorological station, and determine the characteristic relationship of the photovoltaic system by combining the current and voltage parameters output by the photovoltaic array. The output current PV of the photovoltaic array is related to the light intensity S(t), the photovoltaic array temperature and the internal impedance parameters of the circuit, which can be expressed by the following formula: ； Where PV is the output current of the photovoltaic array, U(t) is the output voltage of the photovoltaic array, qPD is the photovoltaic conversion current coefficient, KPD is the photovoltaic output current variation coefficient related to temperature, T is the current temperature of the photovoltaic array, T0 is the reference temperature, Rd and Rc are the series shunt resistance and parallel shunt resistance respectively, ε is the Boltzmann constant, α is the ratio of the shunt resistance to the equivalent internal resistance of the battery, and S(t) is the light intensity; S1.

2. Use voltage and current sensors to obtain the net load change value W(t) of the power grid. The net load of the power grid reflects the dynamic change of the load demand on the system. Its value is affected by the grid voltage, the SOC of the energy storage system and the internal loss of the system. The net load is expressed as: ； Here, W(t) represents the net load of the grid at the current time t, W(t-1) is the net load at the previous time, U(t) and U(t-1) are the grid-connected voltages at the current and previous time, IG(t) and IG(t-1) are the energy storage system SOCs at the current and previous time, G is the total system capacity, and ηg is the grid loss coefficient. This formula can be used to dynamically capture the net load state of the grid at different times. Through the above two formulas, the system can describe the output characteristics of the photovoltaic array and the dynamic changes of the net load of the grid respectively.

3. The method for optimizing hybrid photovoltaic energy storage based on PerturbandObserve algorithm according to claim 2, characterized in that: In S2, it includes: S2.

1. Establish an optimization model based on the characteristics of the photovoltaic energy storage system. The model takes economic optimization as the goal and uses the optimization of control variables to balance the energy interaction between photovoltaic energy storage and the power grid; S2.

2. The model is optimized using the PerturBObserve optimization algorithm. The objective function is the minimum value of the charging and discharging cost at each moment. The economic objective function is expressed as: ； Among them, Fmax is the charging and discharging cost required by the energy storage system when the SOC reaches the maximum value, Fdischarging is the cost of the discharging stage, and Fcharging is the cost of the charging stage. The calculation of the charging and discharging cost needs to consider the efficiency factor, among which the charging efficiency ηc and the discharging efficiency ηd are affected by the internal parameters of the system. The purpose of the objective function is to reduce the overall operating cost by optimizing the charging and discharging process; In S2.3, the SOC constraint model is introduced to limit the SOC range in the optimization process to ensure the safety and stability of system operation. The expression of SOC constraint is: ； Among them, Gm is the actual charge and discharge amount of the system within a day, G is the total system capacity, and IGmax and IGmin are the upper and lower limits of SOC respectively. Through this constraint formula, it is ensured that SOC is always within the safe range during the entire optimization process.

4. The method for optimizing hybrid photovoltaic energy storage based on the PerturbandObserve algorithm according to claim 3 is characterized in that: In S3, it is specifically as follows: S3.

1. Initialize the parameters required by the algorithm: First, randomly determine the decision variables, clarify the optimization objective function. At the same time, set the values of the disturbance pulse width t and the sequence length m in the control strategy, and give the initial state variable x0, such as the initial value of SOC and the current photovoltaic power output. S3.

2. Set the accuracy index εp and the error index εv of the algorithm, and define the random disturbance pulse intensity function f(r), where r is a random variable within the range of [0,1]. This intensity function determines the randomness of the disturbance sequence and affects the convergence speed and result accuracy of the algorithm. S3. Calculate the number of samples and generate the sample set S(x), x ∈ x0, to form the input sample space of the algorithm. The number of samples determines the solution range of the algorithm and the approximation degree to the complex objective function. S3.5-S3.7, generate disturbance sequence and control sequence through algorithm loop variables, disturbance pulse intensity x i The calculation formula is as follows: ； Where b is the offset value of the random variable in the range [-1,1], is the inverse function of the disturbance intensity. The goal of this formula is to generate a random disturbance sequence and explore the neighborhood space of the optimal solution through the disturbance characteristics. On this basis, the control sequence is obtained through weighted calculation: ； Here, w is the control weight coefficient, α is the sequence mean, and K is the optimization sequence length. By controlling the weighted generation of the sequence, the optimal solution is gradually approximated. S3.8 - S3.

9. Evaluate the current control effect by calculating the mean and covariance of the disturbance sequence and the control sequence. The deviation calculation formula is: ， Among them, is the mean of the disturbance sequence, is the mean of the control sequence. The calculation of covariance is used to judge the correlation between the disturbance sequence and the control sequence and guide the optimization direction. When the algorithm loop variable n < Ns, return to S3.6 to continue generating a new disturbance sequence and correcting the control sequence; when the number of samples reaches Ns, judge whether the optimization condition is satisfied according to the accuracy index in S3.

1. If the condition is satisfied, output the final control sequence; otherwise, return to S3.3 to readjust the parameters.

5. The method for optimizing hybrid photovoltaic energy storage based on PerturbandObserve algorithm according to claim 1, characterized in that: It also includes: During the process of optimizing the photovoltaic energy storage system based on the PerturBObserve algorithm, adding the SOC constraint of the energy storage is the key link to achieve the balance between the safety and economy of the system. SOC reflects the current charge and discharge state of the energy storage battery and is an important indicator to measure the available performance of the energy storage device. The SOC constraint can ensure that the energy storage system avoids overcharging and over-discharging during operation, while improving the charge and discharge efficiency and reducing the operation cost. In order to incorporate SOC into the optimization model, it is necessary to introduce a dynamic expression of SOC into the control strategy and define the upper limit of SOC. and lower limit As a constraint condition, considering the effect of the energy storage time interval Δt on the SOC change, the dynamic change of SOC can be described by the following formula: ； in, is the energy storage state at time t+1, is the energy storage state at time t, and are charging power and discharging power respectively, ηc and ηd are charging efficiency and discharging efficiency respectively, is the total capacity of the battery. The goal of this formula is to calculate the change of SOC in real time, combine the power input and system efficiency, and dynamically adjust the operating state of the energy storage system; In the optimization control model, the objective function with SOC as the constraint can be expressed as: ； Among them, F is the total charging and discharging cost in one day, and are the charging and discharging costs in the tth period respectively. The constraints ensure that the SOC operates within a safe range and introduce SOC constraints.

6. The method for optimizing hybrid photovoltaic energy storage based on PerturbandObserve algorithm according to claim 2, characterized in that: In S2, the inputs of the optimization model include the number of time series sampling points T, the variable μt of the optimization control sequence, the mean of the control term, the disturbance sequence length m, the optimization sequence length K, and the correction coefficient σ.

7. The method for optimizing hybrid photovoltaic energy storage based on PerturbandObserve algorithm according to claim 2, characterized in that: In S1.1, the goal of the formula is to deduce the real-time output current of the photovoltaic system based on the known environmental parameters and circuit characteristics, providing an input basis for the scheduling of the energy storage system.

Images