Multi-energy energy management system containing hydrogen energy

Abstract

The invention discloses a multi-energy-source energy management system containing hydrogen energy, and relates to the technical field of electric power, and the multi-energy-source energy management system comprises the steps: collecting the operation data of equipment in a new energy station, and predicting the working state information of the equipment after processing the operation data; combining the working state information of the equipment with task requirements to construct an initial equipment energy flow strategy in the new energy station; and an economic operation optimal objective function is constructed based on the hydrogen production income, and the objective function is solved by using an optimization algorithm to determine an economic optimal energy flow strategy. According to the energy management system provided by the invention, photovoltaic power generation, wind power generation, lithium battery energy storage and a hydrogen fuel cell can be coordinated to participate in frequency modulation and peak regulation of a power grid and hydrogen production by electrolyzing water, energy flow among multiple energies is optimized based on peak and valley electricity prices, and finally, the working state of each energy source is determined by optimizing variables.

Description

Technical Field

[0001] The present invention relates to the field of electric power technology, and in particular to a multi-energy energy management system containing hydrogen energy. Background Art

[0002] With the introduction of the carbon peak and carbon neutrality goals, accelerating the construction of a new power system with a high proportion of renewable energy has become an important task for the development of my country's energy industry. The production cost of renewable energy power generation represented by solar and wind power continues to decline, creating the necessary conditions for the formation of a new power system with renewable energy as the core.

[0003] At the same time, hydrogen energy is an important energy carrier with diverse sources, clean and low-carbon, flexible and efficient characteristics. It is also the main energy source for achieving "carbon peak" and "carbon neutrality". By coordinating and optimizing the working conditions of hydrogen energy and various renewable energy sources, on the one hand, it can help large-scale consumption of renewable energy and avoid the occurrence of large-scale wind and solar power abandonment. On the other hand, it can improve the operating income of new energy stations. However, due to the uneven distribution of resources and the limited capacity of power transmission channels, serious wind and solar power abandonment has occurred in some areas, reducing the economic benefits of energy stations.

[0004] Currently, no effective solution has been proposed for the problems in the related technologies. Summary of the invention

[0005] In order to solve the above problems, the present invention proposes a multi-energy energy management system containing hydrogen energy, which realizes the internal coordinated control of multiple energy sources, thereby improving the economic benefits of new energy stations.

[0006] In order to achieve the above object, the present invention adopts the following technical solution:

[0007] The present invention provides a multi-energy energy management system containing hydrogen energy, the multi-energy energy management system comprising:

[0008] The data acquisition and processing unit is used to collect the operating data of the equipment in the new energy station and predict the working status information of the equipment after processing the operating data;

[0009] Optimizing the scheduling processing unit, which is used to combine the equipment working status information with the task requirements to build the initial equipment energy flow strategy in the new energy station;

[0010] The flow strategy library optimization unit is used to construct the optimal objective function of economic operation based on hydrogen production revenue, and use the optimization algorithm to solve the objective function to determine the economically optimal energy flow strategy.

[0011] Preferably, the data acquisition and processing unit includes:

[0012] Equipment data collection module, used to collect operating data of photovoltaic power generation equipment, wind power generation equipment, lithium battery energy storage equipment, water electrolysis hydrogen production equipment and hydrogen fuel cells in new energy stations;

[0013] The data processing and storage module is used to pre-process the equipment operation data, extract the characteristics of the operation data according to the pre-processing results, and classify and store the operation data based on the characteristic results;

[0014] A power generation curve prediction module is used to input the operating data corresponding to the photovoltaic power generation equipment and the wind power generation equipment as input variables into the recurrent neural network to predict the power generation curves of photovoltaic and wind power;

[0015] The state information prediction module is used to establish state information solution models according to the operating data of lithium battery energy storage, water electrolysis hydrogen production and hydrogen fuel cells, and obtain the working state information of the corresponding equipment.

[0016] Preferably, preprocessing the equipment operation data, extracting the characteristics of the operation data according to the preprocessing results, and implementing classified storage of the operation data based on the characteristic results includes:

[0017] The equipment operation data is sorted and processed in alphabetical units according to the semantic order, and each group of units is divided into data blocks for marking, and then numbers are assigned to the units in each group of data blocks in sequence;

[0018] Compare any two groups of data blocks to see if there is duplicate data. If there is duplicate data, delete the data in the corresponding label. If there is no duplicate data, complete the equipment operation data preprocessing operation;

[0019] The pre-processed equipment operation data is vectorized and converted into a directly readable data language, and the semantic features in the equipment operation data are extracted and fused using feature fusion;

[0020] The semantic features are fused by merging and splicing the output vectors of the paths to obtain a feature fusion vector, and the equipment operation data is classified according to the feature fusion vector;

[0021] Based on the labels of the classified data blocks, the label vectors corresponding to each group of data blocks in the equipment operation data are randomly sorted using a random sorting algorithm, and the data blocks are stored in a set according to the sorting results;

[0022] After using the data encryption algorithm to perform encryption operations on the data in the collection, the encryption result is stored in the index vector to obtain the ciphertext, and the ciphertext is uploaded to the new energy station server to classify and store the operating data.

[0023] Preferably, inputting the operating data corresponding to the photovoltaic power generation equipment and the wind power generation equipment as input variables into the recurrent neural network to predict the photovoltaic and wind power generation curves includes:

[0024] Based on the key decoding ciphertext, the operation data corresponding to the photovoltaic power generation equipment and the wind power generation equipment are extracted from the new energy station server respectively, and the dimension reduction operation is performed on the operation data;

[0025] Obtaining a cost function value of the operating data based on the simplification result, selecting the operating data to perform secondary simplification according to the size of the cost function value, and extracting a feature vector of the operating data using the secondary simplification result;

[0026] After the optimal parameters of the recurrent neural network are generated by the group collaboration algorithm, the operating data feature vectors corresponding to the photovoltaic power generation equipment and the wind power generation equipment are input into the recurrent neural network as input variables;

[0027] The power prediction model corresponding to the photovoltaic and wind power generation equipment is generated based on the recurrent neural network, and the power generation curves of the photovoltaic and wind power generation equipment are obtained according to the power prediction output results.

[0028] Preferably, after the optimal parameters of the recurrent neural network are generated by the group collaboration algorithm, the operating data feature vectors corresponding to the photovoltaic power generation equipment and the wind power generation equipment are input into the recurrent neural network as input variables, including:

[0029] According to the preset maximum number of iterations, the inertia weight coefficient corresponding to the group collaboration algorithm is nonlinearly adjusted, and the fitness value is used to represent the prediction degree value of the recurrent neural network;

[0030] Based on the adjusted group collaboration algorithm, the optimal solution for the predicted degree value is found, and the recurrent neural network space is defined according to the optimal solution to set the feature solution and form a feature population;

[0031] According to the characteristic population, several groups of initial solutions are randomly generated, the optimal values ​​of particles and the optimal values ​​of bee colonies corresponding to the group collaboration algorithm are recorded, and the optimal values ​​of particles and bee colonies corresponding to several groups of initial solutions are compared respectively;

[0032] Based on the comparison result, the maximum value of the particle optimal value and the swarm optimal value is selected as the optimal position of the characteristic solution, and the fitness value of the characteristic solution is determined according to the optimal position;

[0033] According to the fitness value and the optimal position, the Gaussian kernel parameters and regularization coefficients of the recurrent neural network are optimized to generate the best parameter modified recurrent neural network;

[0034] The operating data feature vectors corresponding to the photovoltaic power generation equipment and the wind power generation equipment are input into the modified recurrent neural network as input variables.

[0035] Preferably, the expression of the photovoltaic power generation equipment power prediction model is:

[0036]

[0037] Where P PV Indicates the power generation of photovoltaic power generation equipment, f PV Indicates the derating factor of the photovoltaic array, Y PV represents the photovoltaic array capacity, represents the irradiance reaching the photovoltaic array, represents the irradiance scalar, α P Represents the temperature coefficient of the photovoltaic module, T c Indicates the actual working environment temperature, T c，STC Indicates the temperature under standard test conditions;

[0038] The expression of wind power generation equipment power prediction model is:

[0039]

[0040] Where P wtg (v) represents the power generation capacity of the wind power generation equipment, P r represents the rated power of the fan, v represents the wind speed, and v ci represents the cut-in wind speed, v co Indicates the cut-out wind speed, v r Indicates rated wind speed.

[0041] Preferably, the flow strategy library optimization unit includes:

[0042] The objective function construction module is used to generate the economic optimal objective function according to the operating cost of the equipment in the new energy station and the frequency and peak regulation change state, and obtain the constraint conditions of the economic optimal objective function;

[0043] The objective function solving module is used to solve the economic optimal objective function using the simulated annealing genetic algorithm, and output the working power of the optimal new energy station equipment according to the solution result;

[0044] The energy flow strategy optimization module is used to optimize and adjust the initial equipment energy flow strategy according to the optimal equipment working efficiency, and determine the energy flow strategy with the goal of optimizing economic performance.

[0045] Preferably, the economic optimal objective function is generated according to the operating cost of the equipment in the new energy station and the frequency and peak regulation change state, and the constraint conditions for obtaining the economic optimal objective function include:

[0046] Determine the frequency regulation revenue model based on the daily frequency regulation volume, frequency regulation comprehensive evaluation index and compensation standard, and generate the peak regulation revenue model based on the peak regulation volume and peak regulation price;

[0047] Define the hydrogen production revenue model based on the total amount of hydrogen and the price of hydrogen, and obtain the daily operating costs of photovoltaic power generation equipment, wind power generation equipment, lithium battery energy storage equipment and water electrolysis hydrogen production equipment;

[0048] The daily operation cost, frequency regulation revenue model, peak regulation revenue model, hydrogen production revenue model and access revenue model are integrated to generate the economic optimal objective function;

[0049] Based on the economic optimal objective function, the power demand for frequency and peak regulation and hydrogen production, the power generation power of photovoltaic power generation equipment and wind power generation equipment, and the working efficiency of energy storage batteries and hydrogen fuel cells are obtained as constraints of the economic optimal objective function.

[0050] Preferably, solving the economic optimal objective function using a simulated annealing genetic algorithm, and outputting the optimal working power of the new energy station equipment according to the solution result includes:

[0051] The gene integer permutation coding method is used to represent the working power corresponding to the economic optimum. Each group of genes is set to contain an array with one row and two columns, and the array is used as the column and layer where the device is located for encoding;

[0052] After the encoding is completed, a population with a preset size is randomly generated, and the population is used as the initial population of the economic optimal target search point to obtain the minimum value of the economic optimal target function;

[0053] The reciprocal of the minimum value is taken as the individual fitness value, and the fitness function corresponding to each group of genes is obtained, and then individuals are selected to the initial population according to the probability threshold and the random traversal sampling method;

[0054] After performing crossover and mutation operations on the initial population, a new population is generated, and the new population is used as the operating population of the simulated annealing algorithm. The simulated annealing operation is performed on the operating population to obtain the population group;

[0055] After determining the number of iterations based on the population group, the optimal solution of the economic optimal objective function is output, and the optimal working power of the new energy station equipment is output based on the optimal solution result.

[0056] Preferably, a new population is generated after a crossover mutation operation is performed on the initial population, and the new population is used as an operating population of the simulated annealing algorithm. The population group obtained by performing a simulated annealing operation on the operating population includes:

[0057] Randomly select receptors from the initial population to generate genetic receptors through crossover operation, and perform mutation operation on the genetic receptors to generate a new population and then initialize the control parameters of the simulated annealing algorithm;

[0058] Randomly select a receptor from the new population as the current solution, generate a new solution based on a random perturbation method, determine the probability of the new solution being accepted, and update the new population based on the probability result;

[0059] The temperature is lowered according to the annealing rate and the operations of generating new solutions, accepting probability judgment and updating the population are repeated until the maximum number of iterations is reached to obtain the population group.

[0060] The beneficial effects of the present invention are:

[0061] 1. The energy management system proposed in the present invention can coordinate photovoltaic power generation, wind power generation, lithium battery energy storage, and hydrogen fuel cells to participate in the frequency and peak regulation of the power grid and the electrolysis of water to produce hydrogen, and optimize the energy flow between multiple energy sources based on peak and valley electricity prices, and finally determine the working status of each energy source by optimizing variables.

[0062] 2. The present invention utilizes lithium battery energy storage, the electricity of hydrogen fuel cell system to electrolyze water to produce hydrogen and participates in the frequency and peak regulation of power grid, thereby improving the utilization rate of new energy. At the same time, it combines peak, flat and valley electricity prices, power prediction curves, etc. to coordinate and control multiple energy sources internally, thereby improving the economic benefits of new energy sites.

[0063] 3. The present invention takes the economic operation of the comprehensive hydrogen production system as the optimal objective function, and uses the simulated annealing genetic algorithm to solve and determine the most economically optimal energy flow strategy. It can comprehensively reflect the economy and operation efficiency of the system by comprehensively considering the hydrogen production income, frequency regulation and peak regulation income and energy unit constraints, which helps to ensure the stable operation and safety of the system while pursuing the maximum economic benefit. At the same time, through the probability acceptance mechanism in the simulated annealing process, it avoids the problem of falling into the local optimal solution, and the parallel search characteristics of the genetic algorithm are used to greatly improve the solution efficiency. Therefore, by solving the objective function, the energy flow strategy under the economically optimal conditions can be determined, so that the system can achieve maximum economic benefits while ensuring stable operation and safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0065] Figure 1 is a principle block diagram of a multi-energy energy management system containing hydrogen energy according to an embodiment of the present invention;

[0066] Figure 2 It is a schematic diagram of the connection between a multi-energy energy management system containing hydrogen energy and equipment in a new energy station according to an embodiment of the present invention.

[0067] In the figure:

[0068] 1. Data acquisition and processing unit; 2. Optimization and scheduling processing unit; 3. Flow strategy library optimization unit. DETAILED DESCRIPTION

[0069] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0070] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used in the present invention have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0071] It should be noted that the terms used herein are only for describing specific embodiments, and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that the terms "include" and "have" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0072] In the absence of conflict, the embodiments of the present invention and the features of the embodiments may be combined with each other.

[0073] See also Figure 1 and Figure 2 The present invention provides a multi-energy energy management system containing hydrogen energy, the multi-energy energy management system comprising:

[0074] The data acquisition and processing unit 1 is used to collect the operation data of the equipment in the new energy station, and predict the equipment working status information after processing the operation data;

[0075] The optimization scheduling processing unit 2 is used to combine the equipment working status information with the task requirements to construct the initial equipment energy flow strategy in the new energy station;

[0076] The flow strategy library optimization unit 3 is used to construct the optimal objective function of economic operation based on the hydrogen production income, and use the optimization algorithm to solve the objective function to determine the economic optimal energy flow strategy.

[0077] In this embodiment, the data acquisition and processing unit 1 includes:

[0078] Equipment data collection module, used to collect operating data of photovoltaic power generation equipment, wind power generation equipment, lithium battery energy storage equipment, water electrolysis hydrogen production equipment and hydrogen fuel cells in new energy stations;

[0079] The data processing and storage module is used to pre-process the equipment operation data, extract the characteristics of the operation data according to the pre-processing results, and classify and store the operation data based on the characteristic results;

[0080] A power generation curve prediction module is used to input the operating data corresponding to the photovoltaic power generation equipment and the wind power generation equipment as input variables into the recurrent neural network to predict the power generation curves of photovoltaic and wind power;

[0081] It should be noted that the equipment in new energy stations includes photovoltaic power generation equipment, wind power generation equipment, lithium battery energy storage equipment, water electrolysis hydrogen production equipment and hydrogen fuel cells.

[0082] The state information prediction module is used to establish state information solution models according to the operating data of lithium battery energy storage, water electrolysis hydrogen production and hydrogen fuel cells, and obtain the working state information of the corresponding equipment.

[0083] Specifically, the state information solution model for hydrogen production by water electrolysis is:

[0084] H(t)=η H P ow (t)Δt / (3.6L);

[0085] Where H(t) represents the mass of hydrogen produced during time period t, η H Indicates the equipment conversion efficiency, P ow (t) represents the power input to the hydrogen production equipment during period t, L represents the lower calorific value of hydrogen, and Δt represents the working step.

[0086] When the new energy station receives the frequency modulation instruction, it can release electric energy through the fuel cell to respond to the frequency modulation power, and the fuel cell output voltage calculation formula is:

[0087] U fc =E fc -U loss ;

[0088] Among them, U fc represents the output voltage of the fuel cell, E fc represents the open circuit voltage of the fuel cell, U loss Represents the loss voltage during the electrochemical reaction.

[0089] At the same time, the electrical energy released by the fuel cell is related to the hydrogen consumption rate, which can be expressed as:

[0090]

[0091] Where P fc represents the total power output of the fuel cell, SH represents the hydrogen consumption rate of the fuel cell, M represents the reaction heat of hydrogen, n represents the number of fuel cells in series, and I fcrepresents the output current of the battery cell, and F represents the Faraday constant.

[0092] At the same time, the unit power operating cost of the fuel cell is:

[0093] C hc =C fc +C mt +C sc ;

[0094] In the formula, C hc represents the unit power operating cost of the fuel cell, C fc Represents fuel cost, which is mainly related to factors such as electricity price, equipment cost, and work efficiency. mt It represents the maintenance cost, which is related to the wages of maintenance personnel and the maintenance cost of maintenance equipment. sc Represents the cost of hydrogen storage, which may include the purchase and maintenance costs of hydrogen storage equipment.

[0095] The lithium battery energy storage state information solution model is:

[0096] E0=E m0 -K E (273+T)(1-SOC);

[0097] In the formula, E0 represents the energy storage state of lithium battery, E m0 Indicates the initial voltage of the lithium battery, K E represents the empirical coefficient, T represents the internal temperature of the battery, and SOC represents the current state of charge.

[0098] And the calculation formula of the current state of charge SOC is:

[0099]

[0100] In the formula, Qt represents the energy consumed by the battery, which is the integration of the charge and discharge current i over time t, and Q0 represents the initial capacity of the lithium battery.

[0101] Specifically, when preprocessing the equipment operation data and extracting the features of the operation data according to the preprocessing results, and implementing classified storage of the operation data based on the feature results, the equipment operation data can be sorted and processed in alphabetical units in semantic order, and each group of units can be divided into data blocks for marking operations, and then labels are assigned to the units in each group of data blocks in sequence; any two groups of data blocks are compared to see whether there are duplicate data, and if there are duplicate data, the data in the corresponding labels are deleted, and if there are no duplicate data, the equipment operation data preprocessing operation is completed; the preprocessed equipment operation data is vectorized and converted into a data language that can be directly read, and the equipment operation data is extracted and fused using feature fusion. Semantic features in the operation data; the semantic features are merged and concatenated by the output vector of the path to obtain the feature fusion vector, and the equipment operation data is classified according to the feature fusion vector; based on the label of the classified data block, the label vector corresponding to each group of data blocks in the equipment operation data is randomly sorted by using a random sorting algorithm, and the data blocks are stored in a set according to the sorting result; after the data in the set is encrypted by a data encryption algorithm (asymmetric encryption algorithm, symmetric encryption algorithm and hash algorithm can be used in specific applications), the encryption result is stored in the index vector to obtain the ciphertext, and the ciphertext is uploaded to the new energy station server to classify and store the operation data.

[0102] Specifically, when the operating data corresponding to the photovoltaic power generation equipment and the wind power generation equipment are input as input variables into the recurrent neural network to predict the power generation curves of photovoltaic and wind power, the operating data corresponding to the photovoltaic power generation equipment and the wind power generation equipment can be respectively extracted from the new energy station server based on the key decoding ciphertext, and the operating data can be reduced in dimension; the cost function value of the operating data is obtained based on the reduction result, and the operating data is selected according to the size of the cost function value to perform secondary simplification, and the operating data feature vector is extracted using the secondary simplification result; after the optimal parameters of the recurrent neural network are generated by the group collaboration algorithm, the operating data feature vector corresponding to the photovoltaic power generation equipment and the wind power generation equipment is input as an input variable into the recurrent neural network; the power prediction model corresponding to the photovoltaic wind power generation equipment is generated based on the recurrent neural network, and the power generation curves of the photovoltaic power generation equipment and the wind power generation equipment are obtained according to the power prediction output result.

[0103] Among them, after using the group collaboration algorithm to generate the optimal parameters of the recurrent neural network, when the operating data feature vectors corresponding to the photovoltaic power generation equipment and the wind power generation equipment are input into the recurrent neural network as input variables, the inertia weight coefficient corresponding to the group collaboration algorithm can be nonlinearly adjusted according to the preset maximum number of iterations, and the fitness value is used to represent the prediction degree value of the recurrent neural network; based on the adjusted group collaboration algorithm, the optimal solution of the prediction degree value is found, and the recurrent neural network space is defined according to the optimal solution to set the characteristic solution to form a characteristic population; according to the characteristic population, several groups of initial solutions are randomly generated, and the particle optimal value and the bee colony optimal value corresponding to the group collaboration algorithm are recorded, and the particle optimal value and the bee colony optimal value corresponding to the several groups of initial solutions are compared respectively; based on the comparison result, the maximum value of the particle optimal value and the bee colony optimal value is selected as the optimal position of the characteristic solution, and the fitness value of the characteristic solution is determined according to the optimal position; according to the fitness value and the optimal position, the Gaussian kernel parameter and the regularization coefficient of the recurrent neural network are optimized to generate the optimal parameter to modify the recurrent neural network; the operating data feature vectors corresponding to the photovoltaic power generation equipment and the wind power generation equipment are input into the modified recurrent neural network as input variables.

[0104] It should be explained that the group collaboration algorithm can adopt the artificial bee colony algorithm, and the recurrent neural network can utilize the LSTM neural network algorithm.

[0105] It needs to be explained that the expression of the photovoltaic power generation equipment power prediction model is:

[0106]

[0107] Where P PV Indicates the power generation of photovoltaic power generation equipment, f PV Indicates the derating factor of the photovoltaic array, Y PV represents the photovoltaic array capacity, represents the irradiance reaching the photovoltaic array, represents the irradiance scalar, α P Represents the temperature coefficient of the photovoltaic module, T c Indicates the actual working environment temperature, T c，STC Indicates the temperature under standard test conditions;

[0108] The expression of wind power generation equipment power prediction model is:

[0109]

[0110] Where P wtg (v) represents the power generation capacity of the wind power generation equipment, P r represents the rated power of the fan, v represents the wind speed, and v ci represents the cut-in wind speed, v co Indicates the cut-out wind speed, vr Indicates rated wind speed.

[0111] It should be explained that the initial device energy flow strategy is:

[0112] (1) When there is a demand for frequency regulation and peak regulation, photovoltaic power generation and wind power generation will respond to the required power first. If the power generation does not meet the demand, that is, P demand >P pv +P wtg When necessary, lithium battery storage equipment and hydrogen fuel cells will be coordinated for replenishment;

[0113] P demand =P pv +P wtg +P fc +P bat ;

[0114] Where P demand Represents the power demand, P pv Represents the power of photovoltaic power generation equipment, P wtg Indicates the power of wind power generation equipment, P fc Indicates the power of lithium battery energy storage equipment, P bat Indicates the power of hydrogen fuel cells.

[0115] (2) When wind and solar power are abandoned, this part of the power can be provided to the water electrolysis hydrogen production system. If the power generation is sufficient, that is, P pv +P wtg >P fc , then coordinate the charging of energy storage and fuel cells, that is:

[0116] P pv +P wtg =P fc +P bat ;

[0117] (3) When the hydrogen gap is large, in addition to using photovoltaic power generation and wind power generation to produce hydrogen, the energy storage system can be used to electrolyze water to produce hydrogen, namely:

[0118] P fc =P pv +P wtg +P bat ;

[0119] (4) When there is no special demand and during peak electricity price periods, the energy management system will give priority to the transmission of electricity generated by photovoltaic, wind power, lithium battery and fuel cells from new energy stations, namely:

[0120] P grid =P pv +P wtg +P bat +P fc;

[0121] (5) When there is no special demand and during the off-peak period of electricity prices, the energy management system will give priority to charging energy storage and fuel cells with the power generated by photovoltaic and wind power, and send the excess power to the grid, that is:

[0122] P pv +P wtg =P fc +P bat +P grid ;

[0123] Therefore, the utilization rate of new energy can be improved by utilizing the electricity from "wind and solar power abandonment", lithium battery energy storage, and hydrogen fuel cell systems to electrolyze water to produce hydrogen, as well as participating in grid frequency and peak regulation.

[0124] In this embodiment, the flow strategy library optimization unit 3 includes:

[0125] The objective function construction module is used to generate the economic optimal objective function according to the operating cost of the equipment in the new energy station and the frequency and peak regulation change state, and obtain the constraint conditions of the economic optimal objective function;

[0126] The objective function solving module is used to solve the economic optimal objective function using the simulated annealing genetic algorithm, and output the working power of the optimal new energy station equipment according to the solution result;

[0127] The energy flow strategy optimization module is used to optimize and adjust the initial equipment energy flow strategy according to the optimal equipment working efficiency, and determine the energy flow strategy with the goal of optimizing economic performance.

[0128] Specifically, when generating the economically optimal objective function according to the operating cost of the equipment in the new energy station and the changing state of frequency and peak regulation, and obtaining the constraints of the economically optimal objective function, the frequency regulation benefit model can be determined based on the daily total frequency regulation, the comprehensive frequency regulation evaluation index and the compensation standard, and the peak regulation benefit model can be generated according to the total peak regulation and the peak regulation price; the hydrogen production benefit model is defined based on the total hydrogen and the hydrogen price, and the daily operating costs corresponding to photovoltaic power generation equipment, wind power generation equipment, lithium battery energy storage equipment and water electrolysis hydrogen production equipment are obtained; the daily operating cost, frequency regulation benefit model, peak regulation benefit model, hydrogen production benefit model and online benefit model are integrated to generate the economically optimal objective function; based on the economically optimal objective function, the power demand for frequency regulation and peak regulation and hydrogen production, the power generation power of photovoltaic power generation equipment and wind power generation equipment, and the working efficiency of energy storage batteries and hydrogen fuel cells are obtained as the constraints of the economically optimal objective function.

[0129] It needs to be explained that the expression of the frequency modulation benefit model is:

[0130] C total,1 =P t1 Kp Y AGC ;

[0131] The expression of the peak load shaving revenue model is:

[0132] C total,2 =P t2 C2;

[0133] The expression of hydrogen production revenue model is:

[0134] C total,3 =H t3 C3;

[0135] The expression of the economic optimal objective function is:

[0136] max(C total )=C total,1 +C total,2 +C total,3 +C grid -C PV -C wind -C hc -C BT ;

[0137] In the formula, C total，1 , C total，2 and C total，3 They represent the frequency regulation revenue model, peak regulation revenue model and hydrogen production revenue model respectively, and C gird represents the income from Internet access, C pv , C wind , C hc and C BT Respectively represent the daily operating costs of photovoltaic power generation equipment, wind power generation equipment, water electrolysis equipment and energy storage equipment, K p With Y AGC They represent the comprehensive evaluation index and compensation standard of frequency regulation, C1 and C2 represent the peak regulation price and hydrogen price, respectively. t1 , P t2 , H t3 They respectively represent the total daily frequency regulation, total peak regulation and total hydrogen, and their sizes are all determined by the optimization variables.

[0138] The constraints of the above objective function are:

[0139] Frequency regulation and peak regulation, hydrogen production demand and power supply should always be balanced, that is, meet the following requirements:

[0140] P1(t)+P2(t)+P3(t)=P pv (t)+P grid (t)+P bat (t)+P wtg (t)+Pfs (t);

[0141] In the formula, P1(t), P2(t), and P3(t) represent the power required for frequency regulation, peak regulation, and hydrogen production at time t, respectively. pv (t), P wtg (t) respectively represent the power generation of solar power and wind power at time t, P bat (t), P fs (t) respectively represent the working power of the energy storage battery and the hydrogen fuel cell at time t, P gird (t) represents the grid power.

[0142] The overcharge and over-discharge constraints of energy storage batteries are:

[0143] SOC min <SOC t <SOC max ;

[0144] In the formula, SOC min With SOC max They respectively represent the lower and upper limits of the state of charge of the energy storage battery.

[0145] Among them, when the economic optimal objective function is solved by using the simulated annealing genetic algorithm and the working power of the optimal new energy station equipment is output according to the solution result, the gene integer arrangement coding method can be used to represent the working power corresponding to the economic optimum, and each group of genes is set to contain an array of one row and two columns, and the array is used as the column and layer where the equipment is located for encoding; after the encoding is completed, a population with a preset size is randomly generated, and the population is used as the initial population of the economic optimal target search point to obtain the minimum value of the economic optimal objective function; the reciprocal of the minimum value is used as the individual fitness value, and after obtaining the fitness function corresponding to each group of genes, individuals are selected to the initial population according to the probability threshold and the random traversal sampling method; a new population is generated after a crossover mutation operation is performed on the initial population, and the new population is used as the operating population of the simulated annealing algorithm, and a simulated annealing operation is performed on the operating population to obtain a population group; the optimal solution of the economic optimal objective function is output after judging the number of iterations according to the population group, and the optimal working power of the new energy station equipment is output based on the optimal solution result.

[0146] Specifically, after performing a crossover and mutation operation on the initial population, a new population is generated, and the new population is used as the operating population of the simulated annealing algorithm. When the simulated annealing operation is performed on the operating population to obtain a population group, receptors can be randomly selected from the initial population to generate genetic receptors through a crossover operation, and the genetic receptors are mutated to generate a new population and then the control parameters of the simulated annealing algorithm are initialized; receptors are randomly selected from the new population as the current solution, and a new solution is generated based on a random perturbation method, the probability of the new solution being accepted is judged, and the new population is updated according to the probability result; the temperature is lowered according to the annealing rate and the operations of generating a new solution, judging the acceptance probability and updating the population are repeated until the maximum number of iterations is reached to obtain a population group.

[0147] In summary, with the help of the above technical solutions of the present invention, the energy management system proposed by the present invention can coordinate photovoltaic power generation, wind power generation, lithium battery energy storage, and hydrogen fuel cells to participate in the frequency and peak regulation of the power grid and the electrolysis of water to produce hydrogen, and optimize the energy flow between multiple energies based on the peak and valley electricity prices, and finally determine the working state of each energy source by optimizing the variables. The present invention utilizes the amount of electricity from lithium battery energy storage and hydrogen fuel cell systems to electrolyze water to produce hydrogen and participate in the frequency and peak regulation of the power grid, thereby improving the utilization rate of new energy, and at the same time combines peak and valley electricity prices, power prediction curves, etc. to coordinate and control the internal operation of multiple energy sources, thereby improving the economic benefits of new energy sites.

[0148] The present invention takes the economic operation of the comprehensive hydrogen production system as the optimal objective function, and uses the simulated annealing genetic algorithm to solve and determine the most economically optimal energy flow strategy. It can comprehensively reflect the economy and operation efficiency of the system by comprehensively considering the hydrogen production income, frequency regulation and peak regulation income and energy unit constraints, which helps to ensure the stable operation and safety of the system while pursuing the maximum economic benefit. At the same time, through the probability acceptance mechanism in the simulated annealing process, it avoids the problem of falling into the local optimal solution, and the parallel search characteristics of the genetic algorithm are used to greatly improve the solution efficiency. Therefore, by solving the objective function, the energy flow strategy under the economically optimal conditions can be determined, so that the system can achieve maximum economic benefits while ensuring stable operation and safety.

[0149] Those skilled in the art will appreciate that the units, i.e., algorithm steps, of the various examples described in conjunction with this embodiment can be implemented in electronic hardware or in a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0150] Although the above describes the specific implementation mode of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art on the basis of the technical solution of the present invention without creative work are still within the scope of protection of the present invention.

Claims

1. A multi-energy energy management system containing hydrogen energy, characterized in that: The multi-energy energy management system includes: The data acquisition and processing unit is used to collect the operating data of the equipment in the new energy station and predict the working status information of the equipment after processing the operating data; Optimizing the scheduling processing unit, which is used to combine the equipment working status information with the task requirements to build the initial equipment energy flow strategy in the new energy station; The flow strategy library optimization unit is used to construct the optimal objective function of economic operation based on hydrogen production revenue, and use the optimization algorithm to solve the objective function to determine the economically optimal energy flow strategy.

2. A multi-energy energy management system containing hydrogen energy according to claim 1, characterized in that: The data acquisition and processing unit comprises: Equipment data collection module, used to collect operating data of photovoltaic power generation equipment, wind power generation equipment, lithium battery energy storage equipment, water electrolysis hydrogen production equipment and hydrogen fuel cells in new energy stations; The data processing and storage module is used to pre-process the equipment operation data, extract the characteristics of the operation data according to the pre-processing results, and classify and store the operation data based on the characteristic results; A power generation curve prediction module is used to input the operating data corresponding to the photovoltaic power generation equipment and the wind power generation equipment as input variables into the recurrent neural network to predict the power generation curves of photovoltaic and wind power; The state information prediction module is used to establish state information solution models according to the operating data of lithium battery energy storage, water electrolysis hydrogen production and hydrogen fuel cells, and obtain the working state information of the corresponding equipment.

3. A multi-energy energy management system containing hydrogen energy according to claim 2, characterized in that: The preprocessing of the equipment operation data, extracting the characteristics of the operation data according to the preprocessing result, and implementing classified storage of the operation data based on the characteristic result includes: The equipment operation data is sorted and processed in alphabetical units according to the semantic order, and each group of units is divided into data blocks for marking, and then numbers are assigned to the units in each group of data blocks in sequence; Compare any two groups of data blocks to see if there is duplicate data. If there is duplicate data, delete the data in the corresponding label. If there is no duplicate data, complete the equipment operation data preprocessing operation; The pre-processed equipment operation data is vectorized and converted into a directly readable data language, and the semantic features in the equipment operation data are extracted and fused using feature fusion; The semantic features are fused by merging and splicing the output vectors of the paths to obtain a feature fusion vector, and the equipment operation data is classified according to the feature fusion vector; Based on the labels of the classified data blocks, the label vectors corresponding to each group of data blocks in the equipment operation data are randomly sorted using a random sorting algorithm, and the data blocks are stored in a set according to the sorting results; After using the data encryption algorithm to perform encryption operations on the data in the collection, the encryption result is stored in the index vector to obtain the ciphertext, and the ciphertext is uploaded to the new energy station server to classify and store the operating data.

4. A multi-energy energy management system containing hydrogen energy according to claim 3, characterized in that: The step of inputting the operating data corresponding to the photovoltaic power generation equipment and the wind power generation equipment as input variables into the recurrent neural network to predict the photovoltaic and wind power generation curves includes: Based on the key decoding ciphertext, the operation data corresponding to the photovoltaic power generation equipment and the wind power generation equipment are extracted from the new energy station server respectively, and the dimension reduction operation is performed on the operation data; Obtaining a cost function value of the operating data based on the simplification result, selecting the operating data to perform secondary simplification according to the size of the cost function value, and extracting a feature vector of the operating data using the secondary simplification result; After the optimal parameters of the recurrent neural network are generated by the group collaboration algorithm, the operating data feature vectors corresponding to the photovoltaic power generation equipment and the wind power generation equipment are input into the recurrent neural network as input variables; The power prediction model corresponding to the photovoltaic and wind power generation equipment is generated based on the recurrent neural network, and the power generation curves of the photovoltaic and wind power generation equipment are obtained according to the power prediction output results.

5. A multi-energy energy management system containing hydrogen energy according to claim 4, characterized in that: After the optimal parameters of the recurrent neural network are generated by the group collaboration algorithm, the operating data feature vectors corresponding to the photovoltaic power generation equipment and the wind power generation equipment are input into the recurrent neural network as input variables, including: According to the preset maximum number of iterations, the inertia weight coefficient corresponding to the group collaboration algorithm is nonlinearly adjusted, and the fitness value is used to represent the prediction degree value of the recurrent neural network; Based on the adjusted group collaboration algorithm, the optimal solution for the predicted degree value is found, and the recurrent neural network space is defined according to the optimal solution to set the feature solution and form a feature population; Randomly generate several groups of initial solutions according to the characteristic population, record the optimal values ​​of particles and the optimal values ​​of the swarm corresponding to the group collaboration algorithm, and compare the optimal values ​​of particles and the optimal values ​​of the swarm corresponding to several groups of initial solutions; Based on the comparison result, the maximum value of the particle optimal value and the swarm optimal value is selected as the optimal position of the characteristic solution, and the fitness value of the characteristic solution is determined according to the optimal position; According to the fitness value and the optimal position, the Gaussian kernel parameters and regularization coefficients of the recurrent neural network are optimized to generate the best parameter modified recurrent neural network; The operating data feature vectors corresponding to the photovoltaic power generation equipment and the wind power generation equipment are input into the modified recurrent neural network as input variables.

6. A multi-energy energy management system containing hydrogen energy according to claim 5, characterized in that: The expression of the photovoltaic power generation equipment power prediction model is: Where P PV Indicates the power generation of photovoltaic power generation equipment, f PV Indicates the derating factor of the photovoltaic array, Y PV represents the photovoltaic array capacity, represents the irradiance reaching the photovoltaic array, represents the irradiance scalar, α P Represents the temperature coefficient of the photovoltaic module, T c Indicates the actual working environment temperature, T c，STC Indicates the temperature under standard test conditions; The expression of wind power generation equipment power prediction model is: Where P wtg (v) represents the power generation capacity of the wind power generation equipment, P r represents the rated power of the fan, v represents the wind speed, and v ci represents the cut-in wind speed, v co Indicates the cut-out wind speed, v r Indicates the rated wind speed.

7. A multi-energy energy management system containing hydrogen energy according to claim 1, characterized in that: The flow strategy library optimization unit includes: The objective function construction module is used to generate the economic optimal objective function according to the operating cost of the equipment in the new energy station and the frequency and peak regulation change state, and obtain the constraint conditions of the economic optimal objective function; The objective function solving module is used to solve the economic optimal objective function using the simulated annealing genetic algorithm, and output the working power of the optimal new energy station equipment according to the solution result; The energy flow strategy optimization module is used to optimize and adjust the initial equipment energy flow strategy according to the optimal equipment working efficiency, and determine the energy flow strategy with the goal of optimizing economic performance.

8. A multi-energy energy management system containing hydrogen energy according to claim 7, characterized in that: The generation of the economic optimal objective function according to the operating cost of the equipment in the new energy station and the frequency and peak regulation change state, and obtaining the constraint conditions of the economic optimal objective function include: Determine the frequency regulation revenue model based on the daily frequency regulation volume, frequency regulation comprehensive evaluation index and compensation standard, and generate the peak regulation revenue model based on the peak regulation volume and peak regulation price; Define the hydrogen production revenue model based on the total amount of hydrogen and the price of hydrogen, and obtain the daily operating costs of photovoltaic power generation equipment, wind power generation equipment, lithium battery energy storage equipment and water electrolysis hydrogen production equipment; The daily operation cost, frequency regulation revenue model, peak regulation revenue model, hydrogen production revenue model and access revenue model are integrated to generate the economic optimal objective function; Based on the economic optimal objective function, the power demand for frequency and peak regulation and hydrogen production, the power generation power of photovoltaic power generation equipment and wind power generation equipment, and the working efficiency of energy storage batteries and hydrogen fuel cells are obtained as constraints of the economic optimal objective function.

9. A multi-energy energy management system containing hydrogen energy according to claim 8, characterized in that: The method of solving the economic optimal objective function by using a simulated annealing genetic algorithm and outputting the working power of the optimal new energy station equipment according to the solution result includes: The gene integer permutation coding method is used to represent the working power corresponding to the economic optimum. Each group of genes is set to contain an array with one row and two columns, and the array is used as the column and layer where the device is located for encoding; After the encoding is completed, a population with a preset size is randomly generated, and the population is used as the initial population of the economic optimal target search point to obtain the minimum value of the economic optimal target function; The reciprocal of the minimum value is taken as the individual fitness value, and the fitness function corresponding to each group of genes is obtained, and then individuals are selected to the initial population according to the probability threshold and the random traversal sampling method; After performing crossover and mutation operations on the initial population, a new population is generated, and the new population is used as the operating population of the simulated annealing algorithm. The simulated annealing operation is performed on the operating population to obtain the population group; After determining the number of iterations based on the population group, the optimal solution of the economic optimal objective function is output, and the optimal working power of the new energy station equipment is output based on the optimal solution result.

10. A multi-energy energy management system containing hydrogen energy according to claim 9, characterized in that: The crossover mutation operation is performed on the initial population to generate a new population, and the new population is used as the operating population of the simulated annealing algorithm. The simulated annealing operation is performed on the operating population to obtain a population group including: Randomly select receptors from the initial population to generate genetic receptors through crossover operation, and perform mutation operation on the genetic receptors to generate a new population and then initialize the control parameters of the simulated annealing algorithm; Randomly select a receptor from the new population as the current solution, generate a new solution based on a random perturbation method, determine the probability of the new solution being accepted, and update the new population based on the probability result; The temperature is lowered according to the annealing rate and the operations of generating new solutions, accepting probability judgment and updating the population are repeated until the maximum number of iterations is reached to obtain the population group.

Images