A multi-objective optimization energy management method and device and storage medium

Abstract

The application provides a multi-target optimization energy management method and device and a storage medium, which comprises the following steps: collecting information of a light storage charging station, including a storage charging and discharging power coefficient k of a t time period t , k t ∈[-1, 1], t = 1, 2,..., m, and m represents the number of time periods in a day; establishing an energy conversion system model of the light storage charging station, including a photovoltaic system model, a storage system model, a charging pile system model and an alternating current power distribution network system model; setting multi-dimensional optimization targets, including light storage charging station profit, charging pile satisfaction rate and photovoltaic consumption degree; setting operation constraint conditions of the light storage charging station; solving the multi-target optimization model to obtain an optimal solution of the storage charging and discharging power coefficient k t of the t time period; and forming an optimal energy management strategy according to the mapping of the optimal solution. The application comprehensively optimizes and improves the light storage charging station profit, reduces the degree of power reduction of the charging pile system, and reduces the occurrence of photovoltaic light abandonment.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of optical storage and charging, and particularly relates to a multi-objective optimization energy management method and device and a storage medium. BACKGROUND

[0002] Compared with traditional cars, electric cars have irreplaceable advantages in reducing carbon emissions, reducing the dependence of human society on fossil energy, and reducing traffic pollution, so electric cars will become the mainstream trend. However, with the increase in the number of electric cars, the demand for electric car charging power will be very large, and it has burstiness, randomness and transience in space-time distribution, especially the obvious peak and valley characteristics, which brings great impact on the power grid and affects the safe and reliable operation of the power grid. The optical storage and charging station integrates a photovoltaic system and an energy storage system to ensure the economy and sustainability of energy supply, and also alleviates the impact of electric cars on the power grid. The optimization of the energy management strategy of the optical storage and charging station can better utilize photovoltaic and energy storage to smooth and transfer electricity demand, thereby further alleviating the impact of high electric car penetration on the power grid.

[0003] A good energy management strategy of the optical storage and charging station can greatly utilize photovoltaic and energy storage, and plays a crucial role in the operation optimization of the optical storage and charging station. However, due to unreasonable energy management strategies, the energy storage system fails to reserve sufficient power when the peak period of the charging pile system arrives, and the peak power supply of the alternating power distribution network system is limited, resulting in that the charging pile system must reduce power supply, which seriously affects the charging efficiency of electric cars. In addition, if the energy storage system fails to reserve sufficient capacity, it cannot completely absorb surplus photovoltaic power generation, i.e., photovoltaic light abandonment occurs. Regarding the problems of serious photovoltaic light abandonment and frequent power reduction of the charging pile in the optical storage and charging station, the most relevant research work is the energy management scheme of the optical storage and charging station to maximize profit or minimize cost. These schemes aim to maximize profit or minimize cost, and only solve the problem of operating profit, but they do not effectively solve the problems of serious photovoltaic light abandonment and frequent power reduction of the charging pile. That is, there is no related research work that can propose an energy management scheme that takes both into account. SUMMARY

[0004] The technical problem to be solved by the present application is to provide a multi-objective optimization energy management method and device and a storage medium, which comprehensively optimizes to improve the profit of the optical storage and charging station, reduces the degree of power reduction of the charging pile system, and reduces the occurrence of photovoltaic light abandonment.

[0005] To solve the above technical problems, the technical scheme adopted by the present application is as follows:

[0006] A multi-objective optimization energy management method, comprising the following steps: collecting information of a light storage charging station, including a storage charging and discharging power coefficient k t , k t ∈[-1,1], t=1,2,...,m, m represents the number of time periods in a day; establishing an energy conversion system model of the light storage charging station, including a photovoltaic system model, a storage system model, a charging pile system model and an alternating current power distribution network system model; setting operation constraint conditions of the light storage charging station; setting multi-dimensional optimization objectives, including light storage charging station profit, charging pile satisfaction rate and photovoltaic consumption degree; solving a multi-objective optimization model to obtain an optimal solution of the storage charging and discharging power coefficient k t of the t time period; and forming an optimal energy management strategy according to mapping of the optimal solution.

[0007] A multi-objective optimization energy management system, comprising: a collecting module configured to collect information of a light storage charging station, including a storage charging and discharging power coefficient k t , k t ∈[-1,1], t=1,2,...,m, m represents the number of time periods in a day; a modeling module configured to establish an energy conversion system model of the light storage charging station, including a photovoltaic system model, a storage system model, a charging pile system model and an alternating current power distribution network system model; a condition setting module configured to set operation constraint conditions of the light storage charging station; an objective setting module configured to set multi-dimensional optimization objectives, including light storage charging station profit, charging pile satisfaction rate and photovoltaic consumption degree; a calculation module configured to solve a multi-objective optimization model to obtain an optimal solution of the storage charging and discharging power coefficient k t of the t time period; and a mapping module configured to form an optimal energy management strategy according to mapping of the optimal solution.

[0008] A multi-objective optimization energy management storage medium, the storage medium comprising a stored program, wherein the program executes the above method when running.

[0009] The present application has the following advantages: first, the information of the light storage charging station is collected, and the storage charging and discharging power coefficient k t of the t time period is taken as an optimization variable, then the energy conversion system model of the light storage charging station is established, multi-dimensional optimization objectives are set, and the operation constraint conditions are set based on the energy conversion system model and the multi-dimensional optimization objectives, a self-adaptive energy management scheduling scheme based on multi-objective optimization is proposed, then the optimal solution of k t is obtained by solving the multi-objective optimization model, and finally the optimal solution is mapped to form an optimal energy management strategy. BRIEF DESCRIPTION OF DRAWINGS

[0010] Figure 1 The micro-grid structure diagram of the light storage charging station in the specific embodiment of the present application;

[0011] Figure 2 The overall flowchart for solving the multi-objective optimization model in the specific embodiment of the present application. Specific embodiment

[0012] In order to explain the technical content, the achieved purposes and effects of the present application in detail, the following will be explained in combination with the embodiments and the accompanying drawings.

[0013] The present application takes the light storage charging station of an electric vehicle as the research object, and the micro-grid structure of the light storage charging station is as shown in the figure. Figure 1 The light storage charging station mainly includes the following five system units: a photovoltaic power generation system, an energy storage system, an alternating current power distribution network system, a charging pile system and an energy management system. Among them, the four system units of the photovoltaic power generation system, the energy storage system, the alternating current power distribution network system and the charging pile system complete energy transmission on the direct current bus (DC Bus) through the DC / DC converter or the AC / DC converter. The energy management system generates and issues control instructions to the corresponding DC / DC or AC / DC unit to execute the power regulation of the corresponding module through the multi-objective energy management scheduling strategy, and adjusts the energy transmission power of each system unit on the direct current bus. The present application assumes that the power of each unit is approximately constant within each period, and does not consider the conversion loss of DC / DC and AC / DC and the transmission loss of the line.

[0014] The present application provides a multi-objective optimization energy management method, which comprises the following steps:

[0015] Collecting the information of the light storage charging station, including the energy storage charging and discharging power coefficient k t of the t period, k t ∈[-1,1], t=1,2,...,m, m represents the number of time periods divided in a day;

[0016] Establishing an energy conversion system model of the light storage charging station, including a photovoltaic system model, an energy storage system model, a charging pile system model and an alternating current power distribution network system model;

[0017] Setting the operation constraint condition of the light storage charging station;

[0018] Setting multi-dimensional optimization objectives, including the profit of the light storage charging station, the charging pile satisfaction rate and the photovoltaic accommodation degree;

[0019] Solving the multi-objective optimization model to obtain the optimal solution of the energy storage charging and discharging power coefficient k t of the t period;

[0020] The optimal solution is formed into an optimal energy management strategy according to mapping.

[0021] From the above description, the beneficial effects of the present application are as follows: first, the information of the light storage charging station is collected, the energy storage charging and discharging power coefficient k t is obtained in the t time period, then the energy conversion system model of the light storage charging station is established as an optimization variable, and a multi-dimensional optimization objective is set, and the operation constraint condition is set based on the energy conversion system model and the multi-dimensional optimization objective, an adaptive energy management scheduling scheme based on multi-objective optimization is proposed, then the optimal solution k t is obtained by solving the multi-objective optimization model, and finally the optimal solution is formed into an optimal energy management strategy according to mapping. In this way, an energy management scheduling strategy is proposed, which takes into account improving the profit of the light storage charging station and solving the problems of power reduction of the charging pile and photovoltaic light abandonment.

[0022] Further, the photovoltaic system model is as follows:

[0023]

[0024] wherein, represents the actual consumption power of the photovoltaic system in the t time period, represents the demand power of the charging pile system in the t time period, represents the energy storage charging and discharging power in the t time period, represents the predicted power generation of the photovoltaic system in the t time period;

[0025] The energy storage system model is as follows:

[0026]

[0027] wherein, represents the maximum discharging power of the energy storage, represents the maximum charging power of the energy storage,

[0028]

[0029] wherein, represents the state of charge of the energy storage system at the end of the t time period, and Δt represents the time period length, and respectively represent the lower threshold value and the upper threshold value of the state of charge of the energy storage system, v c is the total capacity of the energy storage,

[0030]

[0031] The charging pile system model is as follows:

[0032]

[0033]

[0034]

[0035] wherein, represents the actual obtained power of the charging pile system in the t period, represents the predicted additional demand power of the charging pile system in the t period, represents the demand power not met by the charging pile system at the end of the t-1 period;

[0036] The AC power distribution network system model is as follows:

[0037]

[0038] wherein, represents the power of the AC power distribution network system connected to the DC bus in the t period.

[0039] From the above description, according to the energy storage charging and discharging power of each period, the energy conversion system model can be used to calculate the energy transmission power of each system unit in each period.

[0040] Further, the profit of the light storage charging station

[0041]

[0042]

[0043] q ess = 2Nd dod v c ,

[0044] wherein, is the charging service unit price of the charging pile system in the t period, is the unit price of purchasing electricity from the AC power distribution network system in the t period, m ess is the degradation loss cost of the energy storage system per unit of electricity, c ess is the purchase cost of the energy storage system, q ess is the full-cycle discharge amount of the energy storage system, N is the full-cycle discharge cycle number, d dod is the discharge depth, β is the energy storage charging and discharging conversion coefficient factor, which is different according to the charging and discharging settings,

[0045]

[0046] wherein, α1 represents the energy storage charging conversion efficiency, and α2 represents the energy storage discharging conversion efficiency;

[0047] The charging pile satisfaction rate

[0048] the photovoltaic accommodation degree

[0049] From the above description, the present application considers the factors of profit, charging pile power supply and photovoltaic light abandonment, and on this basis, an adaptive energy management scheduling scheme is proposed.

[0050] Further, the operation constraint condition of the light storage charging station comprises:

[0051]

[0052] wherein, represents the initial state of charge of the energy storage system, and y represents the cumulative number of time periods.

[0053] From the above description, the energy management scheduling problem to be solved by the present application is a multi-objective optimization problem with constraints.

[0054] Further, the step of solving the multi-objective optimization model to obtain the optimal solution specifically comprises the following steps:

[0055] S1, randomly generate an initial parent population P0, the chromosome of an individual i in the population wherein, represents the energy storage charging and discharging power coefficient of the individual i at the t time period;

[0056] S2, perform crossover and mutation operations on the nth generation parent population P n to generate the nth generation offspring population Q n , n = 0, 1, 2,..., a-1, and a represents the maximum number of iterations;

[0057] S3, combine the nth generation offspring population Q n and the parent population P n into a new population, and perform non-dominant sorting on the new population according to the Pareto dominance relationship, and divide it into multiple non-dominant layers (F1, F2,...);

[0058] S4, construct an (n+1)th generation parent population P n+1 from the first non-dominant layer F1, and gradually add the individuals of the next non-dominant layer to the parent population P n+1 according to the high and low of the Pareto dominance level, until the parent population P n+1 is consistent with the parent population P n in size;

[0059] S5, judging whether the maximum iteration number a is reached, if yes, stopping iteration, and outputting a result, the result being the energy storage charge-discharge power coefficient k of the tth time period t , if not, jumping to the step S2;

[0060] S6, selecting the optimal solution with the minimum weighted sum from the set of Pareto optimal solutions.

[0061] As can be seen from the above description, the non-dominated sorting genetic algorithm is used to solve the multi-objective optimization model, which can maintain the advantages of good individual fitness to obtain more replication opportunities, and also maintains the diversity of the population.

[0062] Further, in the step S4, the last non-dominated layer added is F b , if all the individuals of the non-dominated layer F b are added to the parent population P n+1 , the size of the parent population P n+1 is greater than the parent population P n , and the reference point-based selection mechanism is used to select part of the individuals of the non-dominated layer F b to add to the parent population P n+1 .

[0063] As can be seen from the above description, the reference point-based selection mechanism is used to select the set of Pareto optimal solutions, which can effectively reduce the computational cost and is suitable for processing multiple objectives.

[0064] Further, in the step S2, the individual i uses an adaptive mutation probability

[0065]

[0066] wherein, represents the jth dimension target value of the individual i, represents the minimum jth dimension target value of all individuals of the current population, represents the maximum jth dimension target value of all individuals of the current population, and dim = 1, 2, 3, representing the dimension of the multi-dimensional optimization target.

[0067] Iterating the individual i of the nth generation parent population P n , if the random decimal number , mutation operation is performed, otherwise no mutation operation is performed, and s1 ∈ [0, 1].

[0068] From the above description, the adaptive mutation probability mutation operator can adaptively adjust the mutation probability according to the fitness value of the individual, so that the mutation probability of the individual with poor fitness is large, and the mutation probability of the individual with good fitness is small.

[0069] Further, the fuzzy analytic hierarchy process is used to select the optimal solution with the minimum weighted sum from the Pareto optimal solution set.

[0070] From the above description, the weight of the normalized multi-dimensional fitness value is determined by the FAHP, and the solution with the minimum weighted sum is selected from the Pareto optimal solution set, which can eliminate the subjectivity of the expert system.

[0071] The application also provides a multi-objective optimization energy management system, comprising:

[0072] The acquisition module is used for acquiring information of the light storage charging station, including the energy storage charging and discharging power coefficient k t of the t time period t , k ∈[-1,1], t=1,2,...,m, m represents the number of time periods in a day;

[0073] The modeling module is used for establishing an energy conversion system model of the light storage charging station, including a photovoltaic system model, an energy storage system model, a charging pile system model and an alternating current power distribution network system model;

[0074] The condition setting module is used for setting the operation constraint condition of the light storage charging station;

[0075] The target setting module is used for setting a multi-dimensional optimization target, including the profit of the light storage charging station, the charging pile satisfaction rate and the photovoltaic consumption degree;

[0076] The calculation module is used for solving the multi-objective optimization model to obtain the optimal solution of the energy storage charging and discharging power coefficient k t of the t time period;

[0077] The mapping module is used for forming an optimal energy management strategy according to the mapping of the optimal solution.

[0078] The application also provides a multi-objective optimization energy management storage medium, wherein the storage medium comprises a stored program, and the program performs the above method when running.

[0079] The multi-objective optimization energy management method, device and storage medium are described in the following specific embodiments:

[0080] Embodiment one

[0081] In a first aspect, a multi-objective optimization energy management method comprises the following steps:

[0082] S1, collect information of the light storage charging station, including the energy storage charging and discharging power coefficient k of the t period t , k t ∈[-1, 1], t = 1, 2,..., m, m represents the number of time periods divided in a day. In this embodiment, one day is divided into 48 time periods, that is, m = 48. In this embodiment, the information of the light storage charging station that needs to be collected also includes the charging service price of the charging pile system in each time period, the price of purchasing electricity from the alternating current power distribution network system, the predicted power generation of the photovoltaic system, and the predicted additional demand power of the charging pile system.

[0083] S2, establish an energy conversion system model of the light storage charging station, including a photovoltaic system model, an energy storage system model, a charging pile system model, and an alternating current power distribution network system model.

[0084] Since the power return to the power grid will bring great instability fluctuation to the power grid, and cause great impact on the stable operation of the power grid, the power grid mostly does not accept the return of user power, that is, the excess photovoltaic power generation cannot be returned to the power grid. The power generated by the photovoltaic system, that is, the photovoltaic power generation, is transmitted to the DC bus through the DC-DC converter, and is preferentially supplied to the demand of the charging pile, and the surplus part is stored in the energy storage system. If the energy storage system cannot completely consume the photovoltaic power generation due to capacity limitation, the photovoltaic system will discard the remaining part, that is, the phenomenon of light abandonment occurs.

[0085] 1, the photovoltaic system model is as follows:

[0086] The actual consumption power of the photovoltaic system in the t period can be represented as follows:

[0087]

[0088] wherein, represents the demand power of the charging pile system in the t period, which is determined by the following formula (6), represents the energy storage charging and discharging power in the t period, represents the predicted power generation of the photovoltaic system in the t period, which is not researched in the present application, and is assumed to be known.

[0089] The energy storage system has the functions of digesting the excess photovoltaic power generation, supplementing the insufficient power supply of the charging pile system in the peak period, and reducing the purchase cost of electricity from the alternating current power distribution network system. The energy storage system is connected to the DC bus through the DC / DC converter.

[0090] 2, the energy storage system model is as follows:

[0091] The energy storage charging and discharging power in the t period The variable to be optimized in the application, with a positive value indicating energy storage discharge, a negative value indicating energy storage charging, and the numerical value indicating the size of the charging and discharging power, can be represented as follows:

[0092]

[0093] wherein, represents the maximum discharging power of the energy storage, represents the maximum charging power of the energy storage, k t The optimal solution is obtained by solving a multi-objective optimization model, and the optimization k t is equivalent to optimizing the variable

[0094] represents the state of charge (SOC) of the energy storage system at the end of the tth period, which is updated by subtracting the discharged electric quantity or adding the charged electric quantity from the total capacity v c of the energy storage at the end of the t+1th period, and needs to satisfy the upper and lower threshold limits of SOC, and the update formula is as follows:

[0095]

[0096] wherein, Δt represents the period length, and represent the lower threshold and upper threshold of the state of charge of the energy storage system, respectively, and v c is the total capacity of the energy storage.

[0097] Since the energy storage system has upper and lower threshold limits of SOC, the variable k needs to be adjusted to meet the actual charging and discharging power of the energy storage, and the update formula is as follows:

[0098]

[0099] 3. The charging pile system model is as follows:

[0100] Limited by the supply capacity of the photovoltaic system, the alternating current power distribution network system and the energy storage system, the actual obtained power of the charging pile system in the tth period can be represented as follows:

[0101]

[0102]

[0103]

[0104] wherein, The predicted new demand power of the charging pile system for the tth period is not studied in this paper, and it is assumed to be known, The demand power of the charging pile system for the t-1th period is not met at the end of the period.

[0105] 4. The AC power distribution network system model is as follows:

[0106] The 10kV high voltage AC power of the AC power distribution network system is converted into 380V three-phase AC power through a transformer, and then connected to the DC bus through AC / DC. The power of the AC power distribution network system connected to the DC bus for the tth period can be expressed as follows:

[0107]

[0108] S3, a multi-dimensional optimization target is established, including the profit of the light storage charging station, the charging pile satisfaction rate and the photovoltaic consumption degree.

[0109] The energy management system determines the energy transmission power of each system unit on the DC bus through the multi-objective energy management scheduling strategy. In this embodiment, the optimization target of the multi-objective energy management scheduling strategy is to maximize the profit of the light storage charging station, the charging pile satisfaction rate and the photovoltaic consumption degree.

[0110] The profit r1 of the light storage charging station is obtained by deducting the power cost and the energy storage degradation cost from the turnover, which can be expressed as follows:

[0111]

[0112] wherein, is the charging service unit price of the charging pile system for the tth period, is the unit price of purchasing electricity from the AC power distribution network system for the tth period, m ess is the degradation cost of the energy storage system per unit of electricity, which can be calculated by allocating the purchase cost to the total cycle discharge amount, which can be expressed as follows:

[0113]

[0114] q ess = 2Nd dod v c (11)

[0115] wherein, c ess is the purchase cost of the energy storage system, q ess is the total cycle discharge amount of the energy storage system, N is the total cycle discharge cycle number, and d dod is the discharge depth.

[0116] In addition, β is the energy storage charging and discharging conversion coefficient factor, which is set as different factors according to charging and discharging, which can be expressed as follows:

[0117]

[0118] Wherein, α1 represents the energy storage charging conversion efficiency, and α2 represents the energy storage discharging conversion efficiency.

[0119] The charging pile satisfaction rate r2 is the ratio of the cumulative actual obtained power to the cumulative demand power of the charging pile system in one day, and can be defined as follows:

[0120]

[0121] The photovoltaic consumption degree r3 is the ratio of the cumulative actual consumption power to the cumulative predicted power generation of the photovoltaic system in one day, and can be defined as follows:

[0122]

[0123] In summary, the technical scheme of the present application is: a profit model of a light storage charging station is proposed to calculate the turnover, electricity purchase cost and energy storage degradation loss cost; a charging pile satisfaction rate model is proposed to measure the influence of the power reduction of the charging pile; and a photovoltaic consumption degree model is proposed to measure the influence of photovoltaic light abandonment.

[0124] S4, setting the operation constraint condition of the light storage charging station.

[0125] In the embodiment, the multi-objective energy management scheduling strategy is operated on the basis of the energy conversion system model in step S2 and the multi-dimensional optimization target model in step S3. The corresponding multi-objective energy management scheduling problem is a constrained multi-objective optimization problem, which can be expressed as follows:

[0126]

[0127] Wherein, The initial SOC of the energy storage system is represented by t=0 in the formula, and t=0 is defined as the 0th period, that is, when t=0, y represents the cumulative number of periods, y=1, 2,..., m.

[0128] S5, solving the multi-objective optimization model to obtain the optimal solution of the energy storage charging and discharging power coefficient k t of the tth period.

[0129] 1. Problem coding

[0130] Each individual i is a feasible solution of the multi-objective optimization model, and the chromosome X i of the individual i is as shown in formula (16):

[0131]

[0132] Wherein, Pi,j(t) represents the energy storage charge-discharge power coefficient of individual i in the t time period. That is, each individual in the population is a vector containing the feasible solution information of the multi-objective optimization model. The energy storage charge-discharge power can be calculated by using formula (2), and the energy transmission power of each system unit in each time period can be calculated by using the power relationship of the energy conversion system model.

[0133] 2. Fitness function

[0134] According to the energy transmission power of each system unit in each time period, the target values r1, r2 and r3 of three dimensions can be calculated by using formula (9).

[0135] The jth dimension fitness value of individual i It can be defined as follows:

[0136]

[0137] Wherein, rj(i) represents the jth dimension target value of individual i, rj(min) represents the minimum target value of the jth dimension of all individuals in the current population, rj(max) represents the maximum target value of the jth dimension of all individuals in the current population. The fitness value of each dimension is negatively correlated with its target value. The smaller the fitness value is, the greater the target value represents, and the more ideal individual i is.

[0138] 3. Individual update strategy

[0139] The individual i update steps of the n generation population are as follows. First, the parent population P n and the child population Q n with the population size of c are combined into a new population R n , that is, R n = P n U Q n . From the 2c individuals in the new population R n , c individuals are selected to form the n+1 parent population P n+1 . In order to realize this selection process, the non-dominated sorting of R n is required according to the Pareto dominance relationship, and it is divided into multiple non-dominated layers (F1, F2,...). Then the parent population P n+1 is constructed from the first layer F1, and the individuals of the next layer are gradually added to the parent population P n+1 until its size is equal to c. The last layer of this process is recorded as the bth layer, and all individuals above the b+1th layer are eliminated. If the parent population P n+1The individuals accepted in the last accepted layer (b-th layer) are all removed from the population to keep the size of the population as c. At this time, the diversity of the population is measured to select the individuals to be retained from the b-th layer. In this embodiment, the selection of the individuals in the b-th layer uses the selection mechanism based on the reference point proposed by Deb et al.

[0140] In the basic NSGA-III algorithm, new individuals are generated to construct the offspring population using the crossover operator and the mutation operator. The individuals are crossed according to a fixed probability, and then mutated according to a fixed probability. Due to the indiscriminate mutation mode, i.e., the mutation rates of the better individuals and the worse individuals are consistent, the direction of mutation has randomness and blindness, so it cannot always find the optimal solution. The present application proposes a mutation operator with adaptive mutation probability, which can adaptively adjust the mutation probability according to the fitness value of the individual, so that the mutation probability of the worse individual is large, and the mutation probability of the better individual is small, to solve this problem. The adaptive mutation probability of individual i can be expressed as follows:

[0141]

[0142]

[0143] wherein, represents the adaptive mutation probability of individual i, d i represents the distance between the fitness value of individual i and the ideal point, and dim represents the dimension of the optimization target.

[0144] The polynomial mutation operator of the adaptive mutation probability is shown in Algorithm 1:

[0145]

[0146] The input of Algorithm 1 includes the parent population Pop, the multi-dimensional fitness F = {F1, F2,..., Fdim} of the parent population, and the distribution index η, wherein the multi-dimensional fitness value of individual i is m Firstly, the offspring population is initialized as the parent population (line 1), and the purpose of initialization is to make the code of the individual that has not been mutated consistent with the parent. After initialization, the individuals in the population are traversed, and the mutation probability of individual i is calculated according to F i (line 3), if the random number taking value in [0, 1] is less than the mutation probability of individual i, the mutation operation is performed, otherwise the mutation operation is not performed (line 4).

[0147] Lines 5-20 are the process of performing the mutation operation. Firstly, the individual X after mutation is initialized as the parent individual Pop i ​​​(5th row). Then, the encoding of individual X is traversed, and a mutation operation is performed on each bit using the polynomial mutation operator (7th-19th rows), where s2 is a random number in [0, 1]. h represents the maximum value of the encoding, and l represents the minimum value of the encoding. The meaning of the encoding is to represent the energy storage charging and discharging power coefficient, and therefore its value range is [-1, 1]. Finally, the ith individual Child i is modified into the mutated individual X. After the above mutation process, the final output is the offspring population, which is the Pareto optimal solution set.

[0148] 4. Selecting an optimal solution strategy

[0149] The traditional method for selecting a unique optimal solution from the Pareto optimal solution set generally uses an expert system, which has strong subjectivity. Therefore, the present application uses a fuzzy analytic hierarchy process (FAHP) to eliminate the subjectivity of the expert system. The FAHP method used in the present application is consistent with the method proposed by Qin et al. The present application uses the FAHP method to determine the weight of the normalized multi-dimensional fitness value, so as to select the solution with the minimum weighted sum from the Pareto optimal solution set.

[0150] In summary, the solution of the multi-objective optimization model specifically includes the following steps:

[0151] S51. Initialize the population size, maximum number of iterations and other parameters in the adaptive NSGA-III (AD_NSGA-III) algorithm, create a reference point set, and generate an initial population. The generation rule of the initial population is completely random under the condition of satisfying the operating constraints in step S4.

[0152] wherein the AD_NSGA-III algorithm is used to solve the Pareto optimal solution set of the energy storage charging and discharging power coefficient k t of the tth time period. The algorithm process is shown in Figure 2 .

[0153] S52. According to the individual update strategy, the population is crossed using the binary crossover operator, and the population is mutated using the improved adaptive polynomial mutation operator in Algorithm 1 to generate an offspring population.

[0154] S53. Merge the parent and offspring to form a new population, and perform non-dominated sorting on the new population.

[0155] S54. According to the individual update strategy, the population is standardized, the reference points of the associated individuals are found, and the reference points are used to construct the Pareto optimal solution set.

[0156] S55. Perform the elite retention operation, preserving elite individuals for the next generation. If the maximum number of iterations is reached, output the Pareto optimal solution set; otherwise, jump to step 3.

[0157] S56. Select the optimal solution with the smallest weighted sum from the Pareto optimal solution set according to the FAHP method. The algorithm terminates here.

[0158] S6. The optimal solution is mapped to form the optimal energy management strategy.

[0159] The mapping method from encoded individuals to scheduling results is as follows:

[0160] First, the update is calculated using equations (2)-(4). Equation (2) converts the energy storage charging and discharging power coefficients for each time period into the corresponding energy storage charging and discharging power for that time period. Equations (3) and (4) calculate the corrected energy storage charging and discharging power based on the energy storage capacity limit. Then, using Equations (1), (5), and (8), the actual power absorbed by the photovoltaic system in each time period can be calculated sequentially. Actual power received by the charging pile system at different times Power connected to the DC bus in AC distribution network systems at different times

[0161] The above process yields the power strategy for each system unit in each time period, thus completing the energy management strategy for each time period.

[0162] Secondly, a multi-objective optimization energy management system includes:

[0163] The data acquisition module is used to collect information from the photovoltaic-storage-charging station, including the energy storage charging and discharging power coefficient k in time period t. t k t ∈[-1,1], t=1,2,...,m, where m represents the number of time periods divided into a day. Let k t This serves as an optimization variable in this invention.

[0164] The modeling module is used to establish the energy conversion system model of the photovoltaic-storage charging station, including the photovoltaic system model, energy storage system model, charging pile system model, and AC distribution network system model. Based on the energy storage charging and discharging power at each time period, and using the power relationships in the energy conversion system model, the energy transmission power of each system unit at each time period can be calculated.

[0165] The target setting module is used to set multi-dimensional optimization targets, including the profit of photovoltaic and energy storage charging stations, the charging pile satisfaction rate, and the photovoltaic absorption rate.

[0166] The condition setting module is used to set the operating constraints of the photovoltaic-storage charging station.

[0167] A calculation module is configured to solve the multi-objective optimization model to obtain the optimal solution of the energy storage charge-discharge power coefficient k of the tth time period. t

[0168] The AD_NSGA-III algorithm is used to solve the Pareto optimal solution set of the energy storage charge-discharge power coefficient k of the tth time period. t The AD_NSGA-III algorithm uses a reference point-based method to select the Pareto optimal solution set, and the FAHP method selects the optimal solution with the minimum weighted sum from the Pareto optimal solution set.

[0169] A mapping module is configured to map the optimal solution to form an optimal energy management strategy.

[0170] According to the optimal solution, the energy storage charge-discharge power coefficient k of the tth time period is updated. The power strategy of each system unit of each time period is sequentially calculated, that is, the energy management strategy of each time period is completed.

[0171] In a third aspect, an energy management storage medium for multi-objective optimization includes a stored program, wherein the program performs the above method when executed.

[0172] In summary, the present application provides a multi-objective optimization energy management method, device and storage medium for the energy management and scheduling problem of the photovoltaic storage charging station, optimizes the charge-discharge scheduling strategy of the energy storage system according to the profit of the photovoltaic storage charging station, the charging pile satisfaction rate and the photovoltaic consumption degree, and establishes a multi-objective energy management scheduling strategy. The strategy comprehensively optimizes and improves the profit of the photovoltaic storage charging station, reduces the degree of power supply reduction of the charging pile system, and reduces the occurrence of photovoltaic light abandonment. The energy of the photovoltaic storage charging station can be reasonably managed and scheduled, and the impact of high penetration of electric vehicles on the power grid can be alleviated.

[0173] The above description is only an embodiment of the present application, and does not limit the patent scope of the present application. Any equivalent transformation or direct or indirect application in the related technical field based on the content of the specification and drawings is also included in the patent protection scope of the present application.​

Claims

1. A multi-objective optimization energy management method, characterized by, The method comprises the following steps: Collecting information of the light storage charging station, including the first t Periodic energy storage charging and discharging power coefficient k t , k t ∈[-1,1], t= 1,2,..., m , m Indicates the number of time periods in a day; establishing an energy conversion system model of the light storage charging station, including a photovoltaic system model, a storage system model, a charging pile system model and an alternating current power distribution network system model; setting operation constraint conditions of the light storage charging station; setting multi-dimensional optimization targets, including light storage charging station profit, charging pile satisfaction rate and photovoltaic consumption degree; solving a multi-objective optimization model to obtain the first t energy storage charge and discharge power coefficient of the time period k t optimal solution forming an optimal energy management strategy according to mapping of the optimal solution; the photovoltaic system model is as follows: ， in, Indicates the first t The actual power absorbed by the photovoltaic system during the time period Indicates the first t The power required by the charging station system during certain time periods Indicates the first t Energy storage charging and discharging power during a given period Indicates the first t Predicted power generation of photovoltaic systems during specific time periods; the storage system model is as follows: ， wherein, represents the maximum discharging power of the energy storage, represents the maximum charging power of the energy storage, ， wherein, denotes the t state of charge of the energy storage system at the end of the time period, denotes the length of the time period, and denote a lower threshold value and an upper threshold value, respectively, for the state of charge of the energy storage system, is the total energy storage capacity, ； the charging pile system model is as follows: ， ， ， wherein, denotes the actual obtained power of the charging pile system in the i-th time period, t denotes the actual obtained power of the charging pile system in the i-th time period, denotes the predicted additional demand power of the charging pile system in the i-th time period, t denotes the predicted additional demand power of the charging pile system in the i-th time period, denotes the predicted additional demand power of the charging pile system in the i-th time period, t denotes the demand power not met by the charging pile system at the end of the i-th time period, the alternating current power distribution network system model is as follows: ， in, Indicates the first t The power of the AC distribution network system connected to the DC bus during the specified time period; The light storage charging station profit , ， ， wherein, is the charging service price of the charging pile system in the first time period, t is the charging service price of the charging pile system in the second time period, is the electricity purchase price of the AC power distribution network system in the first time period, t is the electricity purchase price of the AC power distribution network system in the second time period, is the degradation cost of the energy storage system per unit of electricity exchange, is the acquisition cost of the energy storage system, is the full-cycle discharge capacity of the energy storage system, is the full-cycle discharge cycle number, is the depth of discharge, is the energy storage charge and discharge conversion coefficient factor, which is different according to the charge and discharge settings, ， wherein, represents the energy storage charging conversion efficiency, represents the energy storage discharging conversion efficiency; The charging pile satisfaction rate ; The photovoltaic hosting capacity .

2. The multi-objective optimized energy management method of claim 1, wherein: the operation constraint conditions of the light storage charging station include: ， wherein, denotes the initial state of charge of the energy storage system, y denotes the number of accumulated time periods.

3. The multi-objective optimization energy management method according to claim 2, characterized in that: the step of solving the multi-objective optimization model to obtain the optimal solution specifically comprises the following steps: S1, randomly generating an initial parent population chromosomes of individuals in the population i chromosomes of individuals in the population wherein, denotes the energy storage charge-discharge power coefficient of the individual i at the first t time period S2, to the first n Generation of the parent population Crossing and mutation operations are performed to generate the first n Generation of the child population , n= 0, 1, 2,... a -1, a denotes the maximum number of iterations; S3, the first n generation sub-population is combined with the parent population into a new population, and the new population is non-dominantly sorted according to a Pareto dominance relationship, and divided into multiple non-dominant layers ; S4. From the first non-dominated layer S5. Start building a first n +1 generation parent population S6. Add to the parent population one individual from the next non-dominated layer according to the order of the Pareto dominance level until the parent population is the same size as the parent population ; S5, judging whether the maximum iteration number is reached a If yes, stopping iteration, outputting the result, which is the first t period energy storage charging and discharging power coefficient k t Pareto optimal solution set, if not, jumping to step S2; S6, selecting the optimal solution with the minimum weighted sum from the Pareto optimal solution set.

4. The multi-objective optimized energy management method of claim 3, wherein: In the step S4, the last non-dominated layer added is , if all the individuals of the non-dominated layer are added to the parent population , the size of the parent population is greater than the parent population , a selection mechanism based on reference points is used to select part of the individuals of the non-dominated layer to be added to the parent population .

5. The multi-objective optimized energy management method of claim 4, wherein: In said step S2, said individual i Adaptive mutation probability is used , ， wherein, represents the individual i the first j target value of the dimension, represents the minimum target value of the first j dimension of all individuals of the current population, represents the maximum target value of the first j dimension of all individuals of the current population, dim = 1, 2, 3, represents the dimension of the multi-dimensional optimization target; traversing the first n ancestral population of the individual i , if the random number then a mutation operation is performed, otherwise no mutation operation is performed, s 1 ∈ [0, 1].​ 6. The multi-objective optimized energy management method of claim 3, wherein: The optimal solution with the minimum weighted sum is selected from the Pareto optimal solution set by using the fuzzy analytic hierarchy process.

7. A multi-objective optimization energy management system, characterized by, comprise: The collection module is configured to collect information of the light storage charging station, including the first t Energy storage charging and discharging power coefficient of time period k t , k t ∈[-1,1], t= 1,2,..., m , m indicates the number of time periods in a day. a modeling module for establishing an energy conversion system model of the light storage charging station, including a photovoltaic system model, a storage system model, a charging pile system model and an alternating current power distribution network system model; a condition setting module for setting operation constraint conditions of the light storage charging station; a target setting module for setting multi-dimensional optimization targets, including light storage charging station profit, charging pile satisfaction rate and photovoltaic consumption degree; A computing module is configured to solve the multi-objective optimization model to obtain the first t The energy storage charge-discharge power coefficient of the time period k t optimal solution a mapping module for forming an optimal energy management strategy according to mapping of the optimal solution; the photovoltaic system model is as follows: ， wherein, represents the actual curtailed power of the photovoltaic system in the time period, t represents the actual curtailed power of the photovoltaic system in the time period, represents the demand power of the charging pile system in the time period, t represents the demand power of the charging pile system in the time period, represents the charging and discharging power of the energy storage in the time period, t represents the charging and discharging power of the energy storage in the time period, represents the predicted power generation of the photovoltaic system in the time period, t represents the predicted power generation of the photovoltaic system in the time period, the storage system model is as follows: ， wherein, represents the maximum discharging power of the energy storage, represents the maximum charging power of the energy storage, ， wherein, denotes the t state of charge of the energy storage system at the end of the time period, denotes the length of the time period, and denote a lower threshold value and an upper threshold value, respectively, for the state of charge of the energy storage system, is the total energy storage capacity, ； the charging pile system model is as follows: ， ， ， wherein, denotes the actual obtained power of the charging pile system in the i-th time period, t denotes the actual obtained power of the charging pile system in the i-th time period, denotes the predicted additional demand power of the charging pile system in the i-th time period, t denotes the predicted additional demand power of the charging pile system in the i-th time period, denotes the predicted additional demand power of the charging pile system in the i-th time period, t denotes the demand power not met by the charging pile system at the end of the i-th time period. the alternating current power distribution network system model is as follows: ， wherein, represents the power of the AC power distribution grid system accessing the DC bus during the time period; t represents the power of the AC power distribution grid system accessing the DC bus during the time period; The light storage charging station profit , ， ， wherein, is the charging service price of the charging pile system in the first time period, t is the charging service price of the charging pile system in the second time period, is the purchase price of electricity from the AC power grid system in the first time period, t is the purchase price of electricity from the AC power grid system in the second time period, is the degradation cost of the energy storage system per unit of electricity exchanged, is the acquisition cost of the energy storage system, is the full-cycle discharge capacity of the energy storage system, is the full-cycle discharge cycle number of the energy storage system, is the depth of discharge, is the energy storage charge-discharge conversion coefficient factor, which is different according to the charging and discharging settings, ， wherein, represents the energy storage charging conversion efficiency, represents the energy storage discharging conversion efficiency; The charging pile satisfaction rate ; The photovoltaic hosting capacity .

8. A multi-objective optimized energy management storage medium, characterized by, The storage medium comprises a stored program, wherein the program executes the method described in any one of claims 1 to 6 when running.

Images