Hybrid energy storage power optimization method for wind-solar coupling prediction and control system

Abstract

The invention discloses a hybrid energy storage power optimization method for wind-solar coupling prediction and a control system, relates to the field of wind-solar coupling prediction, and solves the problem that most of hybrid energy storage of an existing distributed energy system is static optimization calculation in power distribution research. In order to solve the problems that in the prior art, a wind-light prediction technology is not used for carrying out expected power estimation on an energy storage system, and description of power change characteristics of hybrid energy storage under actual output is lacked, the invention provides the following scheme that the method comprises the following steps: step 1, constructing a multi-scene model for wind-light coupling output prediction; step 2, optimizing and calculating target power of hybrid energy storage under predicted output by PSO (Particle Swarm Optimization); and step 3, a hybrid energy storage power distribution and optimization method. According to the method and the control system, the complexity of processing the constraint conditions in the optimization process can be reduced, and the power characteristics of the whole system and the energy storage units in the hybrid energy storage can be evaluated more simply, conveniently and accurately.

Description

Technical Field

[0001] The present invention relates to the field of wind-solar coupling prediction, and in particular to a hybrid energy storage power optimization method and control system for wind-solar coupling prediction. Background Art

[0002] Wind power generation and photovoltaic power generation are the most representative new energy power generation methods, and their proportion in the energy structure is increasing. Due to the randomness and intermittency of wind and solar energy, the wind-solar coupling power generation method that combines the characteristics of wind power output and photovoltaic output has been greatly developed. On this basis, in order to promote the distributed application of wind-solar coupling systems, it is necessary to consider the uncertainty of adjusting the output of wind-solar coupling systems through energy storage systems after getting rid of large power grids and thermal power units. Given that energy storage includes energy-type energy storage and power-type energy storage, different types of energy storage are generally used to mix to cope with wind and solar output fluctuations at different frequencies. Therefore, how to balance the power distribution of hybrid energy storage has become the main issue to improve system benefits and reduce system costs.

[0003] At present, most of the research on power allocation of hybrid energy storage in distributed energy systems is static optimization calculation, using existing wind and solar data to allocate the power of hybrid energy storage, only focusing on the power situation of hybrid energy storage under recorded data, without using wind and solar prediction technology to estimate the expected power of the energy storage system, and lacking the description of the power change characteristics of hybrid energy storage under actual output. The above factors make it easy for the theoretical calculation results of hybrid energy storage to deviate greatly from the actual output, which has a serious impact on the stability and service life of the system. Therefore, a hybrid energy storage power optimization method and control system based on wind and solar coupling prediction are proposed. Summary of the invention

[0004] The purpose of the present invention is to provide a hybrid energy storage power optimization method and control system with wind-solar coupling prediction, which solves the problem that most of the existing hybrid energy storage research on power distribution of distributed energy systems is static optimization calculation, and does not use wind-solar prediction technology to perform expected power estimation of the energy storage system. At the same time, there is a lack of description of the power change characteristics of the hybrid energy storage under actual output.

[0005] To achieve the above object, the present invention provides the following technical solution: a hybrid energy storage power optimization method based on wind-solar coupling prediction, comprising the following steps:

[0006] Step 1: Construction of a multi-scenario model for wind-solar coupling output prediction: Based on the prediction results, construct multiple scenarios for predicted output;

[0007] Step 2: PSO optimizes and calculates the target power of hybrid energy storage under predicted output: Determine the output power range of hybrid energy storage through multi-scenario results, and optimize the target power of hybrid energy storage;

[0008] Step 3: Power allocation and optimization method of hybrid energy storage: Allocate the power to each energy storage unit in the hybrid energy storage, and optimize the power of each energy storage unit based on the allocation results within the day before.

[0009] Preferably, the construction of the multi-scenario model for wind-solar coupling output prediction in step 1 specifically includes:

[0010] (1) Based on the annual wind power output and photovoltaic output data, the Gumbel-Copula function is used to construct the joint distribution of wind power output and photovoltaic output, and the parameters of the joint distribution are determined using maximum likelihood estimation. The joint distribution function is:

[0011]

[0012] Where: P wind is the annual wind power output data; P pv is the annual photovoltaic output data; ρ is the model parameter;

[0013] (2) Using short-term forecasting technology to predict the wind power output and photovoltaic output data for one day, the records are as follows: and Setting parameters include CDF, M, N, The parameters are:

[0014]

[0015] (3) Using the Monte Carlo MCMC method to perform random sampling, we can obtain the joint distribution random number C of wind power output and photovoltaic output. G-rand , the sampling method is as follows:

[0016] C G-rand = MCMC(CDF,M,N);

[0017] (4) Use the inverse transformation (icdf) of the kernel density estimate (ksdensity) to obtain C G-rand The truth value real of is calculated as follows:

[0018]

[0019] (5) Return the time series of predicted output (sort) to obtain multiple coupled output scenarios for wind and solar output prediction The specific calculation formula is:

[0020]

[0021] Preferably, the target power of the hybrid energy storage under the output predicted by the PSO optimization calculation in step 2 specifically includes:

[0022] (1) Determine the load demand at the load end as P load , calculate the target power of hybrid energy storage, which is different The corresponding power scenario in this case is: The calculation formula is:

[0023]

[0024] (2) Determined in the previous step It is an M×N matrix. According to the characteristics of scenario planning, M∈[2,+∞). Therefore, in the target power scenario undertaken by hybrid energy storage, the maximum possible output of hybrid energy storage can be determined. Minimum possible output Specifically:

[0025]

[0026] (3) By generating a range The random solution in the algorithm is used as the initial particle of the particle swarm algorithm (PSO). The parameters of PSO are set including the number of particles D, particle dimension T, total number of iterations iter, self-learning factor C1, group learning factor C2, initial inertia factor W1, terminal inertia factor W2, speed range [v min , v max ], the particle swarm algorithm optimization calculation is:

[0027] S1: Record the current iteration number as i, the initial number as 1, and randomly generate in the range The initial particles in And in the range [v max , v min ] the initial velocity v(1);

[0028] S2: According to the established model, the objective is to minimize the power shortage rate at the load end under the premise of load demand, and the objective function is set as:

[0029]

[0030] Where: α(i) is the power failure rate at the load end at the i-th iteration;

[0031] S3: Calculate the power failure rate of each particle and determine the particle with the smallest power failure rate and its location and record;

[0032] S4: Particle optimization, when hour, at the same time otherwise and constant;

[0033] S5: Parameter update:

[0034] (a) Update the inertia weight as:

[0035]

[0036] Where: w(i) is the inertia weight at the i-th iteration;

[0037] (b) The update speed is:

[0038]

[0039] (c) Update the particle to:

[0040]

[0041] (d) Out-of-bounds processing of velocity and particles:

[0042] if or but or otherwise No treatment;

[0043] If v(i)v min or v(i)>v max , then v(i)=v min or v(i)=v max , otherwise v(i) is not processed;

[0044] S6: Determine whether the maximum number of iterations has been reached. If so, the optimization ends and the final Otherwise, let i=i+1 and return to step 3 to continue the calculation;

[0045] (4) According to the above method, the target power of hybrid energy storage under the short-term forecast of wind-solar coupling output is finally determined That is, the power that hybrid energy storage is expected to bear on the day before.

[0046] Preferably, the power allocation and optimization method of the hybrid energy storage in step three specifically includes the power allocation of the day-ahead hybrid energy storage and the power optimization adjustment of the intra-day hybrid energy storage.

[0047] Preferably, the power distribution of the day-ahead hybrid energy storage is specifically as follows:

[0048] (1) The target power of hybrid energy storage is allocated by using a sliding average filtering method. The principle of sliding average filtering is as follows:

[0049]

[0050] Where: windowSize is the window width; n is the sampling sequence length;

[0051] The entropy method is used to select the optimal window width. The initial window width is set to T0. The sliding coefficients a and b under this window width are calculated as follows:

[0052]

[0053] Where: ones(1, T0) is a matrix with 1 row and T0 column, and the value is 1;

[0054] When the calculation window width is T0, the target power is allocated to the power P of the battery battery and the power P of the supercapacitor super , specifically:

[0055]

[0056] Where: is a sliding average filter;

[0057] (2) The entropy method is used to calculate the entropy of the battery state of charge (SOC) and the supercapacitor state of charge under different window widths. The window width corresponding to the smallest difference between the battery SOC entropy and the supercapacitor SOC entropy is the optimal window width. The P under the optimal window width is selected. battery and P super As the power consumption of the battery and supercapacitor respectively, the entropy value method is as follows:

[0058]

[0059] Where: X is the SOC value vector with time series; p(x i ) is the probability of the SOC value appearing in the defined interval at the ith moment, which is specifically expressed as follows: In this description, SOC is divided into five intervals, namely 0-0.2, 0.2-0.4, 0.4-0.6, 0.6-0.8, 0.8-1, p(x i ) is the ratio of the number of X appearing in the above intervals to the total time series length;

[0060] Determine the P corresponding to the optimal window width according to the entropy method battery and P super is the target power of each energy storage unit on the previous day.

[0061] Preferably, the power optimization adjustment of the intraday hybrid energy storage is specifically as follows:

[0062] The wind power output and photovoltaic output at a single moment of the day are predicted hour by hour using ultra-short-term prediction technology. Combined with the hourly load demand estimation, the above method is used to calculate the power of the hybrid energy storage at the corresponding moment and the power allocation of each energy storage unit, which is recorded as and The total power of the hybrid energy storage at the corresponding moment is recorded as

[0063] Optimize control based on the target power of hybrid energy storage within the day and the power assumed by each unit energy storage;

[0064] (1) The total power of hybrid energy storage within the day is less than or equal to the target power of hybrid energy storage on the day before, that is,

[0065] when When: Determine the battery power The supercapacitor power is To ensure that the supercapacitor power does not overflow;

[0066] when When: Determine the battery power The supercapacitor power is In this state, the power of each energy storage unit does not need to be adjusted;

[0067] (2) The total power of the intraday hybrid energy storage is greater than the target power of the day-ahead hybrid energy storage, that is,

[0068] Calculating the spillover power of hybrid energy storage The overflow power is allocated using the method described in Section 3 and recorded as and

[0069] judge and P battery The magnitude relationship will affect the excess overflow of a certain energy storage unit power, so the calculation formula for the battery power and supercapacitor power at the corresponding moment is:

[0070]

[0071] A hybrid energy storage control system with wind-solar coupling prediction includes a data acquisition module, a wind-solar power prediction module, a load monitoring module, a power distribution module 1, a power distribution module 2, a power optimization module, and a hybrid energy storage control module.

[0072] Preferably, the data acquisition module is used to collect and store wind and solar resource data and meteorological data, and the relevant data is used for wind and solar power prediction and energy storage power optimization;

[0073] The wind and solar power prediction module is used to perform short-term day-ahead prediction of wind power and photovoltaic power and ultra-short-term hourly prediction within the day, and input multiple scenarios of further prediction results into the power allocation module;

[0074] The load monitoring module is used to monitor the real-time load and provide feedback to the daily hybrid energy storage power optimization;

[0075] The power allocation module 1 is used to calculate the power allocation of the hybrid energy storage on the day before, including a calculation method of the hybrid energy storage target power and a power allocation method of each energy storage unit;

[0076] The power allocation module 2 is used to calculate the power allocation of the hybrid energy storage within a day, including a calculation method of the hybrid energy storage target power and a power allocation method of each energy storage unit;

[0077] The power optimization module is used for power adjustment and optimization of intraday hybrid energy storage;

[0078] The hybrid energy storage control module is used to control the charging and discharging power of the hybrid energy storage according to the results of the power optimization module.

[0079] Compared with the related art, the hybrid energy storage power optimization method and control system provided by the present invention for wind-solar coupling prediction have the following beneficial effects:

[0080] 1. Construct multiple scenarios for prediction results: Generally, prediction results can only provide wind power curves and photovoltaic power curves under a single scenario, and lack descriptions of power changes under wind-solar coupling. At the same time, the uncertainty of wind and solar itself will also have a great impact on the prediction results, further affecting the power changes under wind-solar coupling. The multi-scenario method can include the power changes under wind-solar coupling as much as possible by connecting the correlation of wind-solar output, without considering what factors affect wind-solar power. The more scenarios there are, the more situations are included.

[0081] 2. Target power of energy storage under the predicted output of PSO optimization calculation: Generally, the target power of hybrid energy storage is calculated as the difference between wind and solar output and load demand. In the optimization process, the power, capacity and SOC of multiple hybrid energy storage are required to be constrained, and the optimization steps are increased. Based on the multi-scenario prediction results and load demand, the present invention determines the maximum possible output and minimum possible output range of hybrid energy storage before the optimization calculation begins, that is, the power, capacity and SOC of hybrid energy storage are constrained before the optimization begins, reducing the cumbersomeness of handling constraints during the optimization process.

[0082] 3. Verify the entropy value of hybrid energy storage SOC to determine the power allocation of each energy storage unit: Generally, when performing hybrid energy storage power allocation, no matter which power signal decomposition method is used, when selecting the parameters of signal decomposition, it is often calculated through empirical judgment or gradient descent principle. The calculation process is complicated and the objective function setting lacks universality. The entropy value theory can characterize the degree of disorder of the system. The larger its value, the more disordered the system; the smaller its value, the more ordered the system. It is applied to the power allocation of hybrid energy storage, and the entropy value of each energy storage unit in different SOC intervals is calculated. Figure 6 It can be seen from (a) that the entropy value of the battery SOC changes relatively smoothly, while the entropy value of the supercapacitor SOC changes with certain fluctuations, which is consistent with the power characteristics of the two energy storages, that is, the SOC value of the battery changes smoothly in different SOC intervals, while the SOC value of the supercapacitor changes with greater volatility in different SOC intervals. By selecting the parameters corresponding to the smallest difference in the SOC entropy value of each energy storage unit, the stability between different energy storages will be relatively higher. The entropy theory can more easily and accurately evaluate the power characteristics of the entire system and each energy storage unit within the hybrid energy storage.

[0083] This method and control system can reduce the tediousness of handling constraints during the optimization process, and more easily and accurately evaluate the power characteristics of the entire system and each energy storage unit within the hybrid energy storage. BRIEF DESCRIPTION OF THE DRAWINGS

[0084] Figure 1 This is a block diagram of a hybrid energy storage power optimization method for predicting wind-solar coupling output in multiple scenarios according to the present invention.

[0085] Figure 2 This is a flow chart of the target power of energy storage under the PSO optimization calculation prediction output of the present invention.

[0086] Figure 3 It is a flow chart of the power distribution and optimization method of the hybrid energy storage of the present invention.

[0087] Figure 4 This is a multiple coupled output scenario (10 scenarios) for the wind-solar output prediction of the present invention.

[0088] Figure 5 It is the power shortage rate variation curve in the day-ahead optimization process of the present invention.

[0089] Figure 6 Entropy value verification of the hybrid energy storage SOC of the present invention

[0090] Figure 7 The power changes of each energy storage unit before and during the day after the optimization of the present invention

[0091] Figure 8The power difference between the optimized distributed energy system and the load of the present invention

[0092] Fig. 9 A schematic diagram of a control system for hybrid energy storage in multiple wind-solar coupled output prediction scenarios of the present invention. DETAILED DESCRIPTION

[0093] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments; based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0094] Embodiment 1:

[0095] See also Figure 1-5 The present invention provides a technical solution: a hybrid energy storage power optimization method for wind-solar coupling prediction, comprising the following steps:

[0096] Step 1: Construction of a multi-scenario model for wind-solar coupling output prediction: Based on the prediction results, construct multiple scenarios for predicted output, as shown in the attached manual. Figure 1 As shown, specifically including:

[0097] (1) Based on the annual wind power output and photovoltaic output data, the Gumbel-Copula function is used to construct the joint distribution of wind power output and photovoltaic output. The parameters of the joint distribution are determined using maximum likelihood estimation. The joint distribution function is:

[0098]

[0099] Where: P wind is the annual wind power output data; P pv is the annual photovoltaic output data; ρ is the model parameter;

[0100] (2) Using short-term forecasting technology to predict the wind power output and photovoltaic output data for one day, the records are as follows: and Setting parameters include CDF, M, N, The parameters are:

[0101]

[0102] (3) Using the Monte Carlo MCMC method to perform random sampling, we can obtain the joint distribution random number C of wind power output and photovoltaic output. G-rand , the sampling method is as follows:

[0103] C G-rand = MCMC(CDF,M,N);

[0104] (4) Use the inverse transformation (icdf) of the kernel density estimate (ksdensity) to obtain C G-rand The truth value real of is calculated as:

[0105]

[0106] (5) Return the time series of predicted output (sort) to obtain multiple coupled output scenarios for wind and solar output prediction The specific calculation formula is:

[0107]

[0108] The coupled output multi-scenario results of wind and solar power output prediction are as shown in the attached manual. Figure 4 As shown;

[0109] Step 2: PSO optimizes and calculates the target power of hybrid energy storage under predicted output: Determine the output power range of hybrid energy storage through multi-scenario results, and optimize the target power of hybrid energy storage, as shown in the attached manual. Figure 1 and 2 As shown, specifically including:

[0110] (1) Determine the load demand at the load end as P load , calculate the target power of hybrid energy storage, which is different The corresponding power scenario in this case is: The calculation formula is:

[0111]

[0112] (2) Determined in the previous step It is an M×N matrix. According to the characteristics of scenario planning, M∈[2,+∞). Therefore, in the target power scenario undertaken by hybrid energy storage, the maximum possible output of hybrid energy storage can be determined. Minimum possible output Specifically:

[0113]

[0114] (3) By generating a range The random solution in the algorithm is used as the initial particle of the particle swarm algorithm (PSO). The parameters of PSO are set including the number of particles D, particle dimension T, total number of iterations iter, self-learning factor C1, group learning factor C2, initial inertia factor W1, terminal inertia factor W2, speed range [v min , v max ], the particle swarm algorithm optimization calculation is:

[0115] S1: Record the current iteration number as i, the initial number as 1, and randomly generate in the range The initial particles in And in the range [v max , v min ] the initial velocity v(1);

[0116] S2: According to the established model, the objective is to minimize the power shortage rate at the load end under the premise of load demand, and the objective function is set as:

[0117]

[0118] Where: α(i) is the power failure rate at the load end at the i-th iteration;

[0119] S3: Calculate the power failure rate of each particle and determine the particle with the smallest power failure rate and its location and record;

[0120] S4: Particle optimization, when hour, at the same time otherwise and constant;

[0121] S5: Parameter update:

[0122] (a) Update the inertia weight as:

[0123]

[0124] Where: w(i) is the inertia weight at the i-th iteration;

[0125] (b) The update speed is:

[0126]

[0127] (c) Update the particle to:

[0128]

[0129] (d) Out-of-bounds processing of velocity and particles:

[0130] if or but or otherwise No treatment;

[0131] If v(i) <v min or v(i)>v max , then v(i)=v min or v(i)=vmax , otherwise v(i) is not processed;

[0132] S6: Determine whether the maximum number of iterations has been reached. If so, the optimization ends and the final Otherwise, let i=i+1 and return to step 3 to continue the calculation;

[0133] The optimization results are as shown in the attached manual. Figure 5 As shown: Figure 5 (a) is the sampling interval optimization result; Figure 5 (b) is the optimization result of the effective load interval, that is, the interval in which the load has an effective change. In this embodiment, the interval from 6:00 to 18:00 is selected as the effective load interval.

[0134] (4) According to the above method, the target power of hybrid energy storage under the short-term forecast of wind-solar coupling output is finally determined That is, the power that hybrid energy storage is expected to bear on the day before.

[0135] Step 3: Power allocation and optimization method of hybrid energy storage: allocate the power of each energy storage unit in the hybrid energy storage, and optimize the power of each energy storage unit based on the allocation results of the previous day, as shown in the attached manual. Figure 1 and 3 As shown, it specifically includes the power allocation of the day-ahead hybrid energy storage and the power optimization adjustment of the intraday hybrid energy storage:

[0136] The power distribution of hybrid energy storage is as follows:

[0137] (1) The target power of hybrid energy storage is allocated using the sliding average filtering method. The principle of sliding average filtering is as follows:

[0138]

[0139] Where: windowSize is the window width; n is the sampling sequence length;

[0140] The entropy method is used to select the optimal window width. The initial window width is set to T0. The sliding coefficients a and b under this window width are calculated as follows:

[0141]

[0142] Where: ones(1, T0) is a matrix with 1 row and T0 column, and the value is 1;

[0143] When the calculation window width is T0, the target power is allocated to the power P of the battery battery and the power P of the supercapacitor super , specifically:

[0144]

[0145] Where: is a sliding average filter;

[0146] (2) The entropy method is used to calculate the entropy of the battery state of charge (SOC) and the supercapacitor state of charge under different window widths. The window width corresponding to the smallest difference between the battery SOC entropy and the supercapacitor SOC entropy is the optimal window width. The P under the optimal window width is selected. battery and P super As the power consumption of the battery and supercapacitor respectively, the entropy value method is as follows:

[0147]

[0148] Where: X is the SOC value vector with time series; p(x i ) is the probability of the SOC value appearing in the defined interval at the ith moment, which is specifically expressed as follows: In this description, SOC is divided into five intervals, namely 0-0.2, 0.2-0.4, 0.4-0.6, 0.6-0.8, 0.8-1, p(x i ) is the ratio of the number of X appearing in the above intervals to the total time series length;

[0149] Determine the P corresponding to the optimal window width according to the entropy method battery and P super is the target power of each energy storage unit on the previous day;

[0150] See the entropy value verification of hybrid energy storage SOC in the manual. Figure 6 : Figure 6 (a) is the entropy value of the SOC of each energy storage unit; Figure 6 (b) is the difference between the SOC entropy values ​​of the two energy storages.

[0151] The power optimization adjustment of hybrid energy storage during the day is as follows:

[0152] The ultra-short-term forecasting technology is used to forecast the wind power output and photovoltaic output at a single moment of the day hour by hour. Combined with the hourly load demand estimation, the above method is used to calculate the power of the hybrid energy storage at the corresponding moment and the power allocation of each energy storage unit, which is recorded as and The total power of the hybrid energy storage at the corresponding moment is recorded as

[0153] Optimize control based on the target power of hybrid energy storage within the day and the power assumed by each unit energy storage;

[0154] (1) The total power of hybrid energy storage within the day is less than or equal to the target power of hybrid energy storage on the day before, that is,

[0155] when When: Determine the battery power The supercapacitor power is To ensure that the supercapacitor power does not overflow;

[0156] when When: Determine the battery power The supercapacitor power is In this state, the power of each energy storage unit does not need to be adjusted;

[0157] (2) The total power of the intraday hybrid energy storage is greater than the target power of the day-ahead hybrid energy storage, that is,

[0158] Calculating the spillover power of hybrid energy storage The overflow power is allocated using the method in the third part and recorded as and

[0159] judge and P battery The magnitude relationship will affect the excess overflow of a certain energy storage unit power, so the calculation formula for the battery power and supercapacitor power at the corresponding moment is:

[0160]

[0161] The power changes of each energy storage unit before and during the day after optimization are shown in the attached manual. Figure 7 : Figure 7 (a) The result of the change in the power borne by the battery; Figure 7 (b) The result of power change borne by supercapacitor.

[0162] The difference power between the distributed energy system and the load after optimization is shown in Figure 8 .

[0163] A hybrid energy storage control system with wind-solar coupling prediction includes a data acquisition module, a wind-solar power prediction module, a load monitoring module, a power distribution module 1, a power distribution module 2, a power optimization module, and a hybrid energy storage control module;

[0164] Data acquisition module: used to collect and store wind and solar resource data and meteorological data. The relevant data is used to predict wind and solar power and optimize energy storage power.

[0165] Wind and solar power prediction module: used to make short-term day-ahead predictions of wind power and photovoltaic power and ultra-short-term hourly predictions within the day, and input multiple scenarios of further prediction results into the power allocation module;

[0166] Load monitoring module: used to monitor real-time load and provide feedback to the daily hybrid energy storage power optimization;

[0167] Power allocation module 1: used to calculate the power allocation of the day-ahead hybrid energy storage, including the calculation method of the hybrid energy storage target power and the power allocation method of each energy storage unit;

[0168] Power allocation module 2: used to calculate the power allocation of hybrid energy storage within a day, including the calculation method of the hybrid energy storage target power and the power allocation method of each energy storage unit;

[0169] Power optimization module: used for power adjustment and optimization of intraday hybrid energy storage;

[0170] Hybrid energy storage control module: used to control the charging and discharging power of the hybrid energy storage according to the results of the power optimization module.

Claims

1. A hybrid energy storage power optimization method based on wind-solar coupling prediction, It is characterized in that The following steps are involved: Step 1: Construction of a multi-scenario model for wind-solar coupling output prediction: Based on the prediction results, construct multiple scenarios for predicted output; Step 2: PSO optimizes and calculates the target power of hybrid energy storage under predicted output: Determine the output power range of hybrid energy storage through multi-scenario results, and optimize the target power of hybrid energy storage; Step 3: Power allocation and optimization method of hybrid energy storage: Allocate the power to each energy storage unit in the hybrid energy storage, and optimize the power of each energy storage unit based on the allocation results within the day before.

2. According to claim 1, a hybrid energy storage power optimization method for wind-solar coupling prediction, It is characterized in that The construction of the multi-scenario model for wind-solar coupling output prediction described in step 1 specifically includes: (1) Based on the annual wind power output and photovoltaic output data, the Gumbel-Copula function is used to construct the joint distribution of wind power output and photovoltaic output, and the parameters of the joint distribution are determined using maximum likelihood estimation. The joint distribution function is: Where: P wind is the annual wind power output data; P pv is the annual photovoltaic output data; ρ is the model parameter; (2) Using short-term forecasting technology to predict the wind power output and photovoltaic output data for one day, the records are as follows: and Setting parameters include CDF, M, N, The parameters are: (3) Using the Monte Carlo MCMC method to perform random sampling, we can obtain the joint distribution random number C of wind power output and photovoltaic output. G-rand , the sampling method is as follows: j G-rand =MCMC(CDF,M,N); (4) Use the inverse transformation of kernel density estimation to obtain C G-rand The truth value real of is calculated as follows: (5) Return the time series of predicted output and obtain multiple coupled output scenarios of wind and solar output prediction The specific calculation formula is:

3. According to claim 1, a hybrid energy storage power optimization method for wind-solar coupling prediction, It is characterized in that The target power of the hybrid energy storage under the PSO optimization calculation prediction output described in step 2 specifically includes: (1) Determine the load demand at the load end as P load , calculate the target power of hybrid energy storage, which is different The corresponding power scenario in this case is: The calculation formula is: (2) Determined in the previous step It is an M×N matrix. According to the characteristics of scenario planning, M∈[2,+∞). Therefore, in the target power scenario undertaken by hybrid energy storage, the maximum possible output of hybrid energy storage can be determined. Minimum possible output Specifically: (3) By generating a range The random solution in is used as the initial particle of the particle swarm algorithm. The parameters of the particle swarm algorithm include the number of particles D, particle dimension T, total number of iterations iter, self-learning factor C1, group learning factor C2, initial inertia factor W1, termination inertia factor W2, speed range [v min , v max ], the particle swarm algorithm optimization calculation is: S1: Record the current iteration number as i, the initial number as 1, and randomly generate in the range Initial particles in And in the range [v max , v min ] the initial velocity v(1); S2: According to the established model, the objective is to minimize the power shortage rate at the load end under the premise of load demand, and the objective function is set as: Where: α(i) is the power failure rate at the load end at the i-th iteration; S3: Calculate the power failure rate of each particle and determine the particle with the smallest power failure rate and its location and record; S4: Particle optimization, when hour, at the same time otherwise and constant; S5: Parameter update: (a) Update the inertia weight as: Where: w(i) is the inertia weight at the i-th iteration; (b) The update speed is: (c) Update the particle to: (d) Out-of-bounds processing of velocity and particles: if or but or otherwise No treatment; If v(i) <v min or v(i)>v max , then v(i)=v min or v(i)=v max , otherwise v(i) is not processed; S6: Determine whether the maximum number of iterations has been reached. If so, the optimization ends and the final Otherwise, let i=i+1 and return to step 3 to continue the calculation; (4) According to the above method, the target power of hybrid energy storage under the short-term forecast of wind-solar coupling output is finally determined That is, the power that hybrid energy storage is expected to bear on the day before.

4. According to claim 1, a hybrid energy storage power optimization method for wind-solar coupling prediction, It is characterized in that The power allocation and optimization method of the hybrid energy storage described in step three specifically includes the power allocation of the day-ahead hybrid energy storage and the power optimization adjustment of the intra-day hybrid energy storage.

5. According to claim 4, a hybrid energy storage power optimization method for wind-solar coupling prediction, It is characterized in that The power distribution of the day-ahead hybrid energy storage is specifically as follows: (1) The target power of hybrid energy storage is allocated by using a sliding average filtering method. The principle of sliding average filtering is as follows: Where: windowSize is the window width; n is the sampling sequence length; The entropy method is used to select the optimal window width, and the initial window width is proposed to be T 0 , calculate the sliding coefficients a and b under the window width, the calculation formula is: Where: ones(1, T 0 ) is 1 row T 0 Column, matrix with value 1; Calculate the window width as T 0 When the target power is allocated to the battery power P battery and the power P of the supercapacitor super , specifically: Where: is a sliding average filter; (2) The entropy method is used to calculate the entropy of the battery charge and the supercapacitor charge under different window widths. The window width corresponding to the smallest difference between the battery charge entropy and the supercapacitor charge entropy is the optimal window width. The P under the optimal window width is selected. battery and P super As the power consumption of the battery and supercapacitor respectively, the entropy value method is as follows: Where: X is the charge value vector with time series; p(x i ) is the probability that the battery charge value at the i-th moment appears in the specified interval. Specifically, the battery charge is divided into five intervals, namely 0-0.2, 0.2-0.4, 0.4-0.6, 0.6-0.8, and 0.8-1. p(x i ) is the ratio of the number of X appearing in the above intervals to the total time series length; Determine the P corresponding to the optimal window width according to the entropy method battery and P super is the target power of each energy storage unit on the previous day.

6. A hybrid energy storage power optimization method for wind-solar coupling prediction according to claim 5, It is characterized in that The power optimization adjustment of the intraday hybrid energy storage is specifically as follows: The wind power output and photovoltaic output at a single moment of the day are predicted hour by hour using ultra-short-term prediction technology. Combined with the hourly load demand estimation, the above method is used to calculate the power of the hybrid energy storage at the corresponding moment and the power allocation of each energy storage unit, which is recorded as and The total power of the hybrid energy storage at the corresponding moment is recorded as Optimize control based on the target power of hybrid energy storage within the day and the power assumed by each unit energy storage; (1) The total power of hybrid energy storage within the day is less than or equal to the target power of hybrid energy storage on the day before, that is, when When: Determine the battery power The supercapacitor power is To ensure that the supercapacitor power does not overflow; when When: Determine the battery power The supercapacitor power is In this state, the power of each energy storage unit does not need to be adjusted; (2) The total power of the intraday hybrid energy storage is greater than the target power of the day-ahead hybrid energy storage, that is, Calculating the spillover power of hybrid energy storage The overflow power is allocated using the method described in Section 3 and recorded as and judge and P battery The magnitude relationship will affect the excess overflow of a certain energy storage unit power, so the calculation formula for the battery power and supercapacitor power at the corresponding moment is:

7. A hybrid energy storage control system with wind-solar coupling prediction, It is characterized in that It includes a data acquisition module, a wind and solar power prediction module, a load monitoring module, a power distribution module 1, a power distribution module 2, a power optimization module, and a hybrid energy storage control module.

8. A wind-solar coupling prediction hybrid energy storage control system according to claim 7, Features: The data acquisition module is used to collect and store wind and solar resource data and meteorological data. The relevant data is used for the prediction of wind and solar power and the optimization of energy storage power. The wind and solar power prediction module is used to perform short-term day-ahead prediction of wind power and photovoltaic power and ultra-short-term hourly prediction within the day, and input multiple scenarios of further prediction results into the power allocation module; The load monitoring module is used to monitor the real-time load and provide feedback to the daily hybrid energy storage power optimization; The power allocation module 1 is used to calculate the power allocation of the hybrid energy storage on the day before, including a calculation method of the hybrid energy storage target power and a power allocation method of each energy storage unit; The power allocation module 2 is used to calculate the power allocation of the hybrid energy storage within a day, including a calculation method of the hybrid energy storage target power and a power allocation method of each energy storage unit; The power optimization module is used for power adjustment and optimization of intraday hybrid energy storage; The hybrid energy storage control module is used to control the charging and discharging power of the hybrid energy storage according to the results of the power optimization module.

Images