Quantum-enhanced sarima-based multi-element energy storage configuration method and system for rural micro-grid

Abstract

The application discloses a rural micro-grid multi-element energy storage configuration method and system based on quantum-enhanced SARIMA, and the method comprises the following steps: collecting rural micro-grid operation data, and constructing a multivariate feature set containing new energy consumption indicators and seasonal agricultural tool operation intensity indicators; correcting the candidate parameter space and prediction residual of the SARIMA prediction model; jointly optimizing the non-seasonal parameters and seasonal parameters of the prediction model by using a quantum-enhanced optimization algorithm to obtain the load demand and new energy output prediction results in the target period; establishing a multi-element energy storage configuration optimization model to determine the capacity, power and operation scheduling strategy of electrochemical energy storage, thermal energy storage, hydrogen energy storage, agricultural tool mobile energy storage and adjustable agricultural load. The application combines rural new energy consumption characteristics, agricultural seasonal energy consumption characteristics, quantum-enhanced time series prediction and multi-element energy storage joint configuration, which helps to reduce energy storage redundant configuration and improve the local new energy consumption capacity.

Description

Technical Field

[0001] This invention belongs to the field of rural new energy microgrid operation optimization and energy storage configuration technology, specifically involving a method and system for configuring multi-element energy storage in rural microgrids based on quantum-enhanced SARIMA. Background Technology

[0002] Rural microgrids typically connect to distributed renewable energy sources such as solar, wind, and biomass power, simultaneously supplying electricity for residential use, agricultural production, and public infrastructure. Compared to urban loads, rural microgrid loads exhibit distinct seasonality, cyclicality, and production-driven characteristics. For instance, agricultural production loads, such as irrigation pumping stations, drying equipment, agricultural product cold chain logistics, and agricultural machinery charging / swapping facilities, tend to concentrate during peak periods of sowing, irrigation, harvesting, drying, and cold chain operations, and are influenced by crop planting schedules, weather conditions, agricultural machinery operation plans, and agricultural product processing times.

[0003] Existing microgrid energy storage configuration methods typically allocate capacity based on historical load curves, renewable energy output curves, and economic targets. Some methods introduce time series forecasting models to predict load or renewable energy output. SARIMA (Seasonal Autoregressive Differential Moving Average) can describe time series with seasonality and periodicity, and is suitable for rural microgrid operation data that includes daily, weekly, or agricultural production cycles. However, traditional SARIMA parameter selection usually relies on manual experience, grid search, or local optimization methods. In scenarios with multiple seasons, multiple load types, and multiple renewable energy fluctuation characteristics, it is prone to problems such as low efficiency in parameter combination search, local optima, and large prediction residuals.

[0004] Furthermore, the configuration of energy storage in rural microgrids is not simply a matter of configuring electrochemical energy storage capacity. In agricultural scenarios, thermal energy storage can serve drying, heating, or combined cooling and heating; hydrogen energy storage can absorb surplus electricity from renewable energy sources over extended periods; and mobile energy storage for agricultural machinery and adjustable agricultural loads can participate in peak shaving, emergency power supply, and load shifting during busy farming seasons. If the energy storage configuration model does not consider the seasonal intensity of agricultural operations, the state of renewable energy consumption, and the differentiated response characteristics of various types of energy storage, it can easily lead to redundant energy storage capacity, increased renewable energy curtailment, or insufficient protection of critical agricultural loads.

[0005] Therefore, a technical solution is needed that can combine the characteristics of rural renewable energy consumption, the seasonal energy consumption characteristics of agriculture, the global optimization of time series forecast parameters, and the joint configuration of multiple energy storage systems. Summary of the Invention

[0006] To address the problems existing in the prior art, this invention provides a method and system for configuring multi-element energy storage in rural microgrids based on quantum-enhanced SARIMA.

[0007] This invention adopts the following technical solution: a method for configuring multi-element energy storage in rural microgrids based on quantum-enhanced SARIMA, comprising the following steps:

[0008] S1. Collect rural microgrid operation data, including: new energy power generation data, rural domestic load data, agricultural production load data, seasonal agricultural machinery operation data, meteorological data, agricultural production cycle data, and electricity price data;

[0009] S2. Perform time alignment, outlier processing, missing value correction and normalization on the rural microgrid operation data, and construct a multivariate feature set of the rural microgrid. The multivariate feature set includes at least the renewable energy consumption index and the seasonal agricultural machinery operation intensity index.

[0010] S3. Establish a SARIMA prediction model for predicting rural microgrid load demand and renewable energy output. Based on the renewable energy consumption index and seasonal agricultural machinery operation intensity index, correct the candidate parameter space and prediction residual of the SARIMA prediction model.

[0011] S4. The quantum enhancement optimization algorithm is used to jointly optimize the non-seasonal and seasonal parameters in the SARIMA prediction model to obtain the quantum-enhanced SARIMA prediction model.

[0012] S5. Using the quantum-enhanced SARIMA prediction model, predict the load demand and renewable energy output of the rural microgrid within the target period through rolling correction;

[0013] S6. Based on the predicted load demand and new energy output, establish a multi-element energy storage configuration optimization model, using the rated capacity, rated power and operation scheduling variables of different types of energy storage devices as decision variables;

[0014] S7. Solve the multi-energy storage configuration optimization model and output the multi-energy storage configuration scheme for rural microgrids, including the capacity, power and operation scheduling strategy of different types of energy storage devices within the target period.

[0015] As a preferred embodiment, in step S1, the new energy power generation data includes one or more of the following: photovoltaic power output, wind power output, biomass power generation output, available new energy power output, and curtailed wind and solar power.

[0016] The agricultural production load data includes one or more of the following: irrigation load, drying load, sowing load, harvesting load, agricultural product cold chain load, and agricultural machinery charging and swapping load.

[0017] The meteorological data includes one or more of the following: solar irradiance, temperature, wind speed, humidity, rainfall, and weather type.

[0018] As a preferred option, in step S2, the new energy consumption indicators include one or more of the following: new energy power generation, wind and solar power curtailment, local consumption rate of new energy, fluctuation rate of new energy output, and prediction deviation of new energy output.

[0019] The local consumption rate of new energy and the fluctuation rate of new energy output respectively meet the following requirements:

[0020] ;

[0021] ;

[0022] in, Contribute to the overall development of new energy. Indicates time The local consumption rate of new energy sources Indicates time The local consumption of new energy power. Indicates time New energy power generation and , Indicates time The volatility of new energy power output Indicates the rated installed capacity of new energy sources. Indicates the sampling time interval. This indicates a positive number used to avoid a denominator of zero.

[0023] As a preferred embodiment, in step S2, the seasonal agricultural machinery operation intensity index includes one or more of the following: irrigation load intensity, drying load intensity, sowing load intensity, harvesting load intensity, agricultural product cold chain load intensity, and agricultural machinery charging and swapping load intensity, and is determined in the following manner:

[0024] ;

[0025] ;

[0026] in, Indicates time t Seasonal agricultural machinery operation intensity index r Indicates the type of agricultural production operation. R This indicates the number of types of agricultural production operations. Indicates the first r The weight of agricultural production operations Indicates time t Is it in the first r Indicator variables for agricultural production operations corresponding to the agricultural production cycle. Indicates the first r Agricultural production operations at all times t The load power, Indicates the first r Maximum load power for agricultural production operations Indicates time t The total power of agricultural production load.

[0027] As a preferred embodiment, in step S3, the SARIMA prediction model is a SARIMA model with external adjustment features, and the parameters include non-seasonal parameters. p , d , q and seasonal parameters P , D , Q , s ; in, p , P These represent the non-seasonal autoregressive order and the seasonal autoregressive order, respectively. q , Q These represent the order of the non-seasonal moving average and the order of the seasonal moving average, respectively. d , D Let these represent the non-seasonal difference order and the seasonal difference order, respectively. s Indicates the seasonal cycle; candidate parameter combinations are: .

[0028] The total load demand of the rural microgrid is :

[0029] ;

[0030] The SARIMA prediction model satisfies:

[0031] ;

[0032] in, Indicates the lag operator, Indicates the sequence to be predicted, and Selected from or ; Indicates the rural domestic load power. This indicates the power of loads other than rural domestic loads and agricultural production loads;

[0033] Represents the externally regulated feature vector. Represents meteorological feature vectors. Represents the eigenvector of electricity prices;

[0034] and Let them represent the non-seasonal autoregressive polynomial and the non-seasonal moving average polynomial, respectively. and Let these represent the seasonal autoregressive polynomial and the seasonal moving average polynomial, respectively. Represents the random error term. This represents the correction coefficient of the externally adjusted eigenvector.

[0035] As a preferred embodiment, in step S3, the candidate parameter space of the SARIMA prediction model is corrected based on the new energy consumption index and the seasonal agricultural machinery operation intensity index, including:

[0036] When the volatility of new energy output exceeds the first threshold, the seasonal moving average parameter is increased. Non-seasonal moving average parameters The upper limit of candidates;

[0037] When the seasonal agricultural machinery operation intensity index exceeds the second threshold, the seasonal cycle will be... s The candidate values ​​are set to at least one of the daily cycle, weekly cycle, or agricultural production operation cycle;

[0038] When the local consumption rate of new energy is lower than the third threshold, the weight of the prediction error of new energy output in the objective function of prediction error is increased.

[0039] As a preferred embodiment, in step S4, the quantum enhancement optimization algorithm includes any one of quantum genetic algorithm, quantum particle swarm optimization algorithm, quantum annealing algorithm, or quantum heuristic coding algorithm; the quantum enhancement optimization algorithm encodes the non-seasonal parameters and seasonal parameters with qubit probability amplitude or quantum state probability vector, and obtains candidate parameter combinations through measurement decoding;

[0040] The probability amplitude of the quantum bit is encoded to satisfy:

[0041] ;

[0042] in, Indicates the first Quantum encoding in a generation of quantum populations Indicates the number of bits in the encoding. Indicates the first The generation The rotation angle of a quantum bit. k=1,2,...,K .

[0043] As a preferred embodiment, in step S4, the quantum enhancement optimization algorithm jointly optimizes the non-seasonal and seasonal parameters with the objective of minimizing the prediction error, and the fitness function... satisfy:

[0044] ;

[0045] ;

[0046] ;

[0047] in, This represents the total load demand forecast. This represents the predicted total output of new energy sources. This represents the root mean square error of the total load demand forecast. This represents the root mean square error of the total power output prediction for new energy sources. This represents the mean absolute error of the total load demand forecast. Indicates the number of validation samples. This represents the parameter complexity penalty term. to This represents the weighting coefficient.

[0048] As a preferred embodiment, in step S4, the quantum enhancement optimization algorithm updates the qubit probability amplitude based on the current optimal candidate parameter combination, and the update method satisfies:

[0049] ;

[0050] ;

[0051] in, Indicates the first The generation The rotation angle of a quantum bit. Indicates the increment of the rotation angle. This represents the rotation step size that varies with the number of iterations. This represents the first candidate parameter combination corresponding to the current optimal candidate parameter combination. Each encoding bit Indicates the first value corresponding to the current combination of candidate measurement parameters. Each encoding bit.

[0052] As a preferred embodiment, step S5 also includes a rolling correction step:

[0053] Within the target period, the latest operating data, meteorological data, and agricultural production cycle data are updated according to a preset rolling time interval. Based on the updated multivariate feature set of the rural microgrid, steps S3 to S5 are re-executed to obtain the updated load demand and renewable energy output forecast results.

[0054] As a preferred embodiment, in step S6, the multi-element energy storage includes one or more of the following: lithium battery energy storage, lead-carbon battery energy storage, flow battery energy storage, supercapacitor energy storage, thermal energy storage, hydrogen energy storage, mobile energy storage for agricultural machinery, and adjustable agricultural load.

[0055] The rated capacity, rated power, and operation scheduling variables include at least the first... Rated capacity of energy storage Rated power Charging power Discharge power and energy storage state variables ;

[0056] Among them, when the first When the energy storage is electrochemical energy storage or mobile energy storage for agricultural machinery, It is in a charged state; when the first When the energy storage is thermal energy storage, It is in a heat storage state; when the first When the energy storage is hydrogen energy storage, This refers to the hydrogen storage state or equivalent energy state.

[0057] As a preferred embodiment, in step S6, the objective function of the multi-element energy storage configuration optimization model satisfies:

[0058] ;

[0059] in, This represents the overall configuration target value. Indicates the investment cost of energy storage. Indicates operating and maintenance costs. This indicates the cost of purchasing and selling electricity. This indicates the penalty cost for abandoning wind and solar power. Indicates the cost of energy storage losses. This indicates the cost of power supply reliability penalties. This represents the penalty cost for the peak-to-valley difference.

[0060] As a preferred embodiment, the energy storage investment cost, operation and maintenance cost, electricity purchase and sale cost, wind and solar curtailment penalty cost, energy storage loss cost, power supply reliability penalty cost, and peak-valley difference penalty cost respectively satisfy the following:

[0061] ;

[0062] ;

[0063] ;

[0064] ;

[0065] ;

[0066] ;

[0067] ;

[0068] in, Represents a set of energy storage types. Indicates the first The unit capacity cost of energy storage Indicates the first The unit power cost of energy storage Indicates the first Class of energy storage rated capacity, Indicates the first Class of energy storage rated power, Indicates the conversion factor; Indicates the first The unit operation and maintenance cost of energy storage systems; and Representing time respectively The purchase price and the sale price of electricity. and These represent the power purchased from the grid and the power sold, respectively. This represents the penalty coefficient for wind and solar power curtailment. Indicates the amount of wind and solar power curtailed; Indicates the first The unit charge / discharge loss cost of energy storage devices This represents the power supply reliability penalty coefficient. This indicates that the load power requirement has not been met. This represents the peak-to-valley difference penalty coefficient.

[0069] As a preferred embodiment, the constraints of the multi-energy storage configuration optimization model include power balance constraints, energy storage state of charge constraints, energy storage capacity constraints, energy storage charging and discharging power constraints, thermal energy storage energy balance constraints, hydrogen energy storage quality balance constraints, agricultural machinery mobile energy storage access time constraints, adjustable agricultural load operation time window constraints, and power supply reliability constraints.

[0070] The power balance constraint satisfies:

[0071] ;

[0072] The electrochemical energy storage state of charge constraint satisfies:

[0073] ;

[0074] ;

[0075] The energy balance constraint of the thermal energy storage satisfies:

[0076] ;

[0077] ;

[0078] The hydrogen energy storage mass balance constraint satisfies:

[0079] ;

[0080] ;

[0081] The adjustable agricultural load operation time window constraint satisfies:

[0082] ;

[0083] ;

[0084] in, This represents the predicted total power output of the new energy source obtained from the quantum-enhanced SARIMA prediction model. This represents the total load demand forecast obtained from the quantum-enhanced SARIMA prediction model. This indicates that the load power requirement has not been met. and They represent the first The discharge power and charging power of energy storage devices Indicates the state of charge of electrochemical energy storage. Indicates the rated capacity of electrochemical energy storage. Indicates the thermal energy storage state. Indicates the rated capacity of thermal energy storage. Indicates hydrogen storage capacity. Indicates hydrogen storage capacity. Indicates the equivalent energy capacity of hydrogen storage. Indicates the hydrogen production capacity by electrolysis. Indicates the power generation capacity of the fuel cell. This indicates the lower heating value of hydrogen. and They represent the first The earliest start time and the latest finish time for each agricultural task. and They represent the first The actual start and end times of each agricultural operation task. Indicates the first The electricity required for each agricultural operation task.

[0085] As a preferred option, in step S7, the annual operating cycle is divided into the off-season, sowing season, irrigation season, harvesting season, drying season or cold chain peak season according to the agricultural production season, and the corresponding energy storage operation scheduling strategy is determined for each.

[0086] Specifically, during the irrigation period, the available capacity constraint weight of electrochemical energy storage and mobile energy storage of agricultural machinery is increased; during the drying period, the coordinated scheduling weight of thermal energy storage and biomass power generation is increased; during the peak cold chain period, the power supply reliability constraint weight is increased; and during the off-season, the energy storage charge-discharge cycle depth is reduced to reduce energy storage lifespan loss.

[0087] The present invention also provides: a multi-element energy storage configuration system for rural microgrids based on quantum-enhanced SARIMA, used to implement the aforementioned method, comprising:

[0088] The data acquisition module is used to collect rural microgrid operation data, including new energy power generation data, rural domestic load data, agricultural production load data, seasonal agricultural machinery operation data, meteorological data, agricultural production cycle data, and electricity price data.

[0089] The feature construction module is used to perform time alignment, outlier processing, missing value correction and normalization on the rural microgrid operation data, and to construct a multivariate feature set of the rural microgrid that includes new energy consumption index and seasonal agricultural machinery operation intensity index.

[0090] The quantum-enhanced prediction module is used to establish a SARIMA prediction model. Based on the multivariate feature set of the rural microgrid, the candidate parameter space and prediction residual of the SARIMA prediction model are corrected. The quantum-enhanced optimization algorithm is used to jointly optimize the non-seasonal and seasonal parameters in the prediction model to obtain the load demand and new energy output prediction results within the target period.

[0091] The energy storage configuration optimization module is used to establish and solve a multi-element energy storage configuration optimization model based on the load demand and new energy output forecast results, so as to determine the capacity, power and operation scheduling strategy of different types of energy storage devices.

[0092] The solution output module is used to output the multi-energy storage configuration solution for rural microgrids.

[0093] Compared with the prior art, the present invention, employing the above technical solution, has the following technical effects:

[0094] 1. This invention introduces the renewable energy consumption index and the seasonal agricultural machinery operation intensity index into the SARIMA parameter search and prediction residual correction, which helps to improve the adaptability of rural microgrid load and renewable energy output prediction to agricultural seasonal fluctuations.

[0095] 2. This invention uses a quantum-enhanced optimization algorithm to jointly optimize both non-seasonal and seasonal parameters of SARIMA, which helps to reduce the impact of manual parameter setting and local optima on the prediction results.

[0096] 3. This invention inputs the prediction results into a multi-element energy storage configuration optimization model, which helps to reduce the redundancy of a single energy storage configuration and enhances the complementary regulation capabilities among electrochemical energy storage, thermal energy storage, hydrogen energy storage, mobile energy storage for agricultural machinery, and adjustable agricultural loads.

[0097] 4. The present invention sets up differentiated scheduling strategies according to the off-season, sowing season, irrigation season, harvesting season, drying season and cold chain peak season, which helps to improve the ability to guarantee key agricultural loads and the ability to consume new energy locally. Attached Figure Description

[0098] Figure 1 This is a flowchart of the multi-energy storage configuration method for rural microgrids according to the present invention;

[0099] Figure 2 This is a historical simulation data curve of the rural microgrid in the embodiment;

[0100] Figure 3 This is a convergence curve of quantum-enhanced SARIMA parameter optimization in the embodiment;

[0101] Figure 4 This is a comparison curve of the target period prediction results in the example;

[0102] Figure 5 This is a bar chart showing the results of the multi-energy storage capacity configuration in the embodiments;

[0103] Figure 6 The diagram shows the operation and scheduling curves of the multi-energy storage system in this embodiment. Detailed Implementation

[0104] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of the application will be further described in detail below with reference to the accompanying drawings. The described embodiments are only a part of the embodiments involved in this invention. All non-innovative embodiments based on these embodiments by other researchers in the art are within the protection scope of this invention. Furthermore, the step numbers in the embodiments of this invention are only set for ease of explanation and do not limit the order of the steps. The execution order of each step in the embodiments can be adaptively adjusted according to the understanding of those skilled in the art.

[0105] This invention discloses a method for configuring multi-element energy storage in rural microgrids based on quantum-enhanced SARIMA, the process of which is as follows: Figure 1 As shown, it includes the following steps:

[0106] S1. Collect rural microgrid operation data, which includes new energy power generation data, rural domestic load data, agricultural production load data, seasonal agricultural machinery operation data, meteorological data, agricultural production cycle data, and electricity price data.

[0107] S2. Perform time alignment, outlier processing, missing value correction and normalization on the rural microgrid operation data, and construct a multivariate feature set of the rural microgrid. The multivariate feature set of the rural microgrid includes at least the new energy consumption index and the seasonal agricultural machinery operation intensity index.

[0108] S3. Establish a SARIMA prediction model for predicting rural microgrid load demand and renewable energy output, and correct the candidate parameter space and prediction residual of the SARIMA prediction model based on the renewable energy consumption index and the seasonal agricultural machinery operation intensity index.

[0109] S4. The quantum enhancement optimization algorithm is used to jointly optimize the non-seasonal and seasonal parameters in the SARIMA prediction model to obtain the quantum-enhanced SARIMA prediction model.

[0110] S5. Use the quantum-enhanced SARIMA prediction model to predict the load demand and renewable energy output of the rural microgrid within the target period;

[0111] S6. Based on the predicted load demand and new energy output, establish a multi-electrode energy storage configuration optimization model, wherein the rated capacity, rated power and operation scheduling variables of different types of energy storage devices are used as decision variables.

[0112] S7. Solve the multi-energy storage configuration optimization model and output the multi-energy storage configuration scheme for rural microgrids. The multi-energy storage configuration scheme for rural microgrids includes the capacity, power and operation scheduling strategy of different types of energy storage devices within the target period.

[0113] In one embodiment of the present invention, a multi-energy storage configuration is performed on a rural microgrid using the method of the present invention.

[0114] The rural microgrid in this embodiment includes photovoltaic power generation units, wind power generation units, biomass power generation units, grid connection interfaces, lithium battery energy storage, thermal energy storage, hydrogen energy storage, mobile energy storage for agricultural machinery, and several adjustable agricultural loads. Agricultural production loads include irrigation pump station loads, grain drying loads, agricultural product cold chain loads, and agricultural machinery charging and swapping loads. The sampling time interval is set to 1 hour, and the target period is set to the next week.

[0115] In this embodiment, the SARIMA model is used to describe time series with seasonality or periodicity.

[0116] Quantum enhancement optimization algorithms are optimization algorithms that use the ideas of quantum bit probability amplitude, quantum state probability vector, quantum rotation gate or quantum annealing for global search. They include quantum genetic algorithm, quantum particle swarm optimization algorithm, quantum annealing algorithm and quantum heuristic coding algorithm.

[0117] The renewable energy consumption index is a set of indicators used to characterize the renewable energy generation, local consumption, wind and solar curtailment, and fluctuation of rural microgrids.

[0118] The seasonal agricultural machinery operation intensity index is used to characterize the energy intensity of agricultural production operations such as irrigation, sowing, harvesting, drying, cold chain, and agricultural machinery charging and swapping at different times.

[0119] Multi-element energy storage refers to one or more of the following energy storage or equivalent energy storage resources that work together to participate in the capacity configuration and operation scheduling of microgrids: electrochemical energy storage, thermal energy storage, hydrogen energy storage, mobile energy storage for agricultural machinery, and adjustable agricultural loads.

[0120] The target period is used to predict and configure future time windows for optimization, which can be the current 24 hours, the next week, a typical season, or a yearly operating period.

[0121] Step 1: Perform data collection.

[0122] Data collected includes historical 90-day data on photovoltaic power output, wind power output, biomass energy output, rural domestic load, irrigation load, drying load, cold chain load, agricultural machinery charging and swapping load, meteorological data, agricultural production cycle data, and time-of-use electricity price data. The agricultural production cycle includes the off-season, irrigation season, drying season, and peak cold chain season.

[0123] The historical simulation data in this embodiment includes the total load demand curve, the total output curve of new energy sources, and the seasonal agricultural machinery operation intensity curve, such as... Figure 2 As shown, this represents the relationship between total load demand, total output of new energy sources, and seasonal agricultural machinery operation intensity over time.

[0124] Total load demand is formed by the superposition of rural living load, agricultural production load and other loads. Total new energy output is formed by the superposition of photovoltaic output, wind power output and biomass energy output. Seasonal agricultural machinery operation intensity is used to reflect the impact of agricultural production operations such as irrigation, drying, cold chain and agricultural machinery charging and swapping on load patterns at different times.

[0125] Step two: Perform data preprocessing and feature construction.

[0126] The data obtained in step one is then subjected to time alignment, outlier handling, missing value correction, and normalization to obtain a data sequence with a uniform time scale.

[0127] Then, the renewable energy consumption index and the seasonal agricultural machinery operation intensity index are calculated, and the meteorological characteristics, agricultural production cycle characteristics and electricity price characteristics are combined into a multivariate characteristic set of rural microgrids.

[0128] Step 3: Establish and refine the SARIMA prediction model.

[0129] Using the total load of rural microgrids and the total output of renewable energy as the prediction targets, a SARIMA prediction model with external regulation characteristics is established. When the volatility of renewable energy output is high, the upper limit of the candidate parameters for moving averages is increased; when the seasonal intensity of agricultural machinery operation is high, the agricultural production operation cycle is included in the seasonal cycle candidate set; when the local consumption rate of renewable energy is low, the weight of the renewable energy output prediction error in the fitness function is increased.

[0130] Step four: Perform quantum enhancement parameter optimization.

[0131] A quantum-enhanced optimization algorithm is employed to jointly optimize the non-seasonal and seasonal parameters of the SARIMA prediction model. In each iteration, the quantum code is first measured and decoded to obtain candidate parameter combinations; then, the fitness of the candidate parameter combinations is calculated using training and validation data; finally, the quantum code is updated based on the current optimal candidate parameter combination. After the iteration terminates, the optimal parameter combinations corresponding to the load prediction model and the new energy output prediction model are obtained.

[0132] During the parameter optimization process, the fitness of the load demand prediction model and the renewable energy output prediction model is updated with the number of iterations. In this embodiment, the convergence curve of the quantum-enhanced SARIMA parameter optimization is as follows: Figure 3 The figure shows the fitness changes of the load demand forecasting model and the renewable energy output forecasting model during the iteration process, illustrating the process of the quantum augmentation optimization algorithm performing a global search and iterative update of candidate parameter combinations. The decrease in fitness indicates a combined reduction in the prediction error and complexity penalty corresponding to the candidate SARIMA parameter combinations.

[0133] Step 5: Perform target cycle prediction.

[0134] Using the quantum-enhanced SARIMA prediction model obtained in step four, the total load and total output of new energy sources in the rural microgrid for the coming week are predicted, resulting in load demand prediction curves and new energy output prediction curves for energy storage configuration optimization. If new meteorological data, load data, or agricultural production plan data are obtained during operation, rolling adjustments are made.

[0135] The target cycle forecast results include the actual total load demand, the predicted total load demand, the actual total output of renewable energy, and the predicted total output of renewable energy. The predicted total load demand is denoted as... The predicted total output of new energy sources is denoted as Both serve as inputs to the subsequent multi-element energy storage configuration optimization model.

[0136] In this embodiment, the target period prediction result comparison curve is as follows: Figure 4The diagram illustrates the correspondence between the actual value of total load demand, the predicted value of total load demand, the actual value of total renewable energy output, and the predicted value of total renewable energy output.

[0137] Step 6: Establish and solve the multi-element energy storage configuration optimization model.

[0138] Based on the prediction curves obtained in step five, a multi-element energy storage configuration optimization model is established. This model uses electrochemical energy storage capacity and power, thermal energy storage capacity and power, hydrogen energy storage capacity and power, mobile energy storage capacity and power for agricultural machinery, and adjustable agricultural load operation time windows as decision variables. It incorporates investment cost, operation and maintenance cost, electricity purchase and sale cost, wind and solar curtailment penalty cost, energy storage loss cost, power supply reliability penalty cost, and peak-valley difference penalty cost as comprehensive objectives, while also satisfying constraints such as power balance, energy storage status, agricultural operation time windows, and power supply reliability.

[0139] Step 7: Output the multi-energy storage configuration scheme and seasonal operation strategy.

[0140] Based on the solution results from step six, output the capacity, power, charging and discharging plans, and seasonal scheduling strategies for different types of energy storage devices.

[0141] Seasonal operation strategies include: prioritizing the availability of mobile energy storage capacity for irrigation pump stations and agricultural machinery during the irrigation season; prioritizing the coordination of thermal energy storage and biomass power generation during the drying season; increasing the weight of power supply reliability constraints during peak cold chain periods; and reducing the energy storage cycle depth during the off-season to minimize energy storage lifespan loss.

[0142] In a simulated data-based running example, the optimal parameters for load forecasting output by the MATLAB implementation program are: The optimal parameters for new energy prediction are: .

[0143] Given the simulation cost parameters and operational constraints, the output example configuration results are as follows: electrochemical energy storage capacity 480.0 kWh, power 216.0 kW; thermal energy storage capacity 80.0 kWh, power 28.0 kW; hydrogen energy storage capacity 400.0 kWh, power 100.0 kW; mobile energy storage capacity for agricultural machinery 180.0 kWh, power 99.0 kW; and a comprehensive target value of 7042.91.

[0144] The above results are only used to illustrate the output format and calculation process of the embodiments of the present invention, and are not intended to limit the actual engineering effect.

[0145] The multi-energy storage capacity configuration results output in this embodiment include recommended capacities for electrochemical energy storage, thermal energy storage, hydrogen energy storage, and mobile energy storage for agricultural machinery, such as... Figure 5 As shown in the figure, this diagram illustrates that multiple types of energy storage resources are not configured individually, but are jointly determined based on prediction curves and constraints.

[0146] The execution scheduling results output by the example are as follows: Figure 6 As shown, the charge / discharge power of electrochemical energy storage, mobile energy storage for agricultural machinery, and hydrogen energy storage within the target cycle is as follows: Figure 6 (a) in the figure, and the corresponding energy storage state trajectory, such as Figure 6 (b) In this context, a positive charge / discharge power indicates discharging or supplying energy, while a negative charge / discharge power indicates charging or absorbing surplus renewable energy. The energy storage state trajectory is used to illustrate that different energy storage systems meet capacity and state constraints within the target period.

[0147] In some implementations, the quantum-enhanced optimization algorithm can be replaced by a quantum genetic algorithm, a quantum particle swarm optimization algorithm, or a quantum annealing algorithm. The SARIMA prediction model can also be extended to SARIMAX, decomposed SARIMA, or a multi-model ensemble prediction structure; where SARIMA is used to characterize periodic trends, and other models are used to characterize nonlinear residuals.

[0148] In some implementations, multi-element energy storage may further include supercapacitors, flywheel energy storage, pumped hydro storage, compressed air energy storage, cold storage energy storage, and electric agricultural machinery fleets. The multi-element energy storage configuration optimization model can be solved using mixed-integer linear programming, quadratic programming, dynamic programming, genetic algorithms, particle swarm optimization, or rolling optimization algorithms.

[0149] In some implementations, the division of agricultural production cycles can be adjusted according to different crop types in different regions. For example, rice-growing areas can focus on the transplanting period, irrigation period, harvesting period, and drying period; facility agriculture areas can focus on the greenhouse supplemental lighting period, heating period, and peak cold chain period; and livestock breeding areas can focus on the peak periods of ventilation, heat preservation, feeding, and cold chain load.

[0150] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for configuring multi-element energy storage in rural microgrids based on quantum-enhanced SARIMA, characterized in that, Includes the following steps: S1. Collect rural microgrid operation data, including: new energy power generation data, rural domestic load data, agricultural production load data, seasonal agricultural machinery operation data, meteorological data, agricultural production cycle data, and electricity price data; S2. Perform time alignment, outlier processing, missing value correction and normalization on the rural microgrid operation data, and construct a multivariate feature set of the rural microgrid. The multivariate feature set includes at least the renewable energy consumption index and the seasonal agricultural machinery operation intensity index. S3. Establish a SARIMA prediction model for predicting rural microgrid load demand and renewable energy output. Based on the renewable energy consumption index and seasonal agricultural machinery operation intensity index, correct the candidate parameter space and prediction residual of the SARIMA prediction model. S4. The quantum enhancement optimization algorithm is used to jointly optimize the non-seasonal and seasonal parameters in the SARIMA prediction model to obtain the quantum-enhanced SARIMA prediction model. S5. Using the quantum-enhanced SARIMA prediction model, predict the load demand and renewable energy output of the rural microgrid within the target period through rolling correction; S6. Based on the predicted load demand and new energy output, establish a multi-element energy storage configuration optimization model, using the rated capacity, rated power and operation scheduling variables of different types of energy storage devices as decision variables; S7. Solve the multi-energy storage configuration optimization model and output the multi-energy storage configuration scheme for rural microgrids, including the capacity, power and operation scheduling strategy of different types of energy storage devices within the target period.

2. The method for configuring multi-source energy storage in rural microgrids according to claim 1, characterized in that, In step S1, the new energy power generation data includes one or more of the following: photovoltaic power output, wind power output, biomass power generation output, available new energy power output, and curtailed wind and solar power output; the agricultural production load data includes one or more of the following: irrigation load, drying load, sowing load, harvesting load, agricultural product cold chain load, and agricultural machinery charging and swapping load; the meteorological data includes one or more of the following: solar irradiance, temperature, wind speed, humidity, rainfall, and weather type.

3. The method for configuring multi-energy storage in rural microgrids according to claim 1, characterized in that, In step S2, the new energy consumption indicators include one or more of the following: new energy power generation, wind and solar power curtailment, local consumption rate of new energy, fluctuation rate of new energy output, and prediction deviation of new energy output. The local consumption rate of new energy and the fluctuation rate of new energy output respectively meet the following requirements: ； ； in, This represents the local consumption rate of new energy at time t. This represents the amount of electricity consumed locally by new energy sources at time t. This represents the amount of new energy power generation at time t. This represents the fluctuation rate of new energy power output at time t. Indicates the rated installed capacity of new energy sources. This represents the total output of new energy sources at time t. Indicates the sampling time interval. This indicates a positive number used to avoid a denominator of zero.

4. The method for configuring multi-energy storage in rural microgrids according to claim 3, characterized in that, In step S2, the seasonal agricultural machinery operation intensity index includes one or more of the following: irrigation load intensity, drying load intensity, sowing load intensity, harvesting load intensity, agricultural product cold chain load intensity, and agricultural machinery charging and swapping load intensity, and is determined in the following manner: ； ； in, Indicator of seasonal agricultural machinery operation intensity at time t. Indicates the type of agricultural production operation. This indicates the number of types of agricultural production operations. Indicates the first The weight of agricultural production operations Indicates whether time t is the first... Indicator variables for agricultural production operations corresponding to the agricultural production cycle. Indicates the first The load power of agricultural production operations at time t. Indicates the first Maximum load power for agricultural production operations This represents the total agricultural production load at time t.

5. The method for configuring multi-source energy storage in rural microgrids according to claim 4, characterized in that, In step S3, the SARIMA prediction model is a SARIMA model with external adjustment features, and the parameters include non-seasonal parameters. p , d , q and seasonal parameters P , D , Q , s ; in, p , P These represent the non-seasonal autoregressive order and the seasonal autoregressive order, respectively. q , Q These represent the order of the non-seasonal moving average and the order of the seasonal moving average, respectively. d , D Let these represent the non-seasonal difference order and the seasonal difference order, respectively. s Indicates seasonal cycle; candidate parameter combination ; The total load demand of the rural microgrid for: ； in, Indicates the rural domestic load power. This indicates the power of loads other than rural domestic loads and agricultural production loads; The SARIMA prediction model satisfies: ； in, Indicates the lag operator; The sequence to be predicted is selected from... or ; Represents the externally regulated feature vector. and Let them represent the non-seasonal autoregressive polynomial and the non-seasonal moving average polynomial, respectively. and Let these represent the seasonal autoregressive polynomial and the seasonal moving average polynomial, respectively. Represents the random error term. Represents the correction coefficient of the externally adjusted eigenvector; Based on the aforementioned new energy consumption index and seasonal agricultural machinery operation intensity index, the candidate parameter space of the SARIMA prediction model is corrected: When the volatility of new energy output exceeds the first threshold, the seasonal moving average parameter is increased. Non-seasonal moving average parameters The upper limit of candidates; When the seasonal agricultural machinery operation intensity index exceeds the second threshold, the seasonal cycle will be... s The candidate values ​​are set to at least one of the daily cycle, weekly cycle, or agricultural production operation cycle; When the local consumption rate of new energy is lower than the third threshold, the weight of the prediction error of new energy output in the objective function of prediction error is increased.

6. The method for configuring multi-energy storage in rural microgrids according to claim 5, characterized in that, In step S4, the quantum enhancement optimization algorithm encodes the non-seasonal and seasonal parameters with qubit probability amplitude or quantum state probability vector, and obtains candidate parameter combinations through measurement decoding, including any one of quantum genetic algorithm, quantum particle swarm algorithm, quantum annealing algorithm or quantum heuristic encoding algorithm; The probability amplitude of the quantum bit is encoded to satisfy: ； in, Indicates the first Quantum encoding in a generation of quantum populations Indicates the number of bits in the encoding. Indicates the first The generation The rotation angle of a quantum bit. k=1,2,...,K .

7. The method for configuring multi-source energy storage in rural microgrids according to claim 6, characterized in that, In step S4, the quantum enhancement optimization algorithm jointly optimizes the non-seasonal and seasonal parameters with the goal of minimizing the prediction error, and the fitness function... satisfy: ； ； ； in, This represents the total load demand forecast. This represents the predicted total output of new energy sources. This represents the root mean square error of the total load demand forecast. This represents the root mean square error of the total power output prediction for new energy sources. This represents the mean absolute error of the total load demand forecast. Indicates the number of validation samples. This represents the parameter complexity penalty term. to Indicates the weighting coefficient; The quantum enhancement optimization algorithm updates the qubit probability amplitude based on the current optimal candidate parameter combination, and the update method satisfies: ； ； in, Indicates the first The generation The rotation angle of a quantum bit. Indicates the increment of the rotation angle. This represents the rotation step size that varies with the number of iterations. This represents the first candidate parameter combination corresponding to the current optimal candidate parameter combination. Each encoding bit Indicates the first value corresponding to the current combination of candidate measurement parameters. Each encoding bit.

8. The method for configuring multi-energy storage in rural microgrids according to claim 7, characterized in that, In step S6, the multi-element energy storage includes one or more of the following: lithium battery energy storage, lead-carbon battery energy storage, flow battery energy storage, supercapacitor energy storage, thermal energy storage, hydrogen energy storage, mobile energy storage for agricultural machinery, and adjustable agricultural load. The rated capacity, rated power, and operational scheduling variables include at least the first... Rated capacity of energy storage Rated power Charging power Discharge power and energy storage state variables Among them, when the first When the energy storage is electrochemical energy storage or mobile energy storage for agricultural machinery, It is in a charged state; when the first When the energy storage is thermal energy storage, It is in a heat storage state; when the first When the energy storage is hydrogen energy storage, This refers to the hydrogen storage state or equivalent energy state. The objective function of the multi-element energy storage configuration optimization model satisfies: ； in, This represents the overall configuration target value. Indicates the investment cost of energy storage. Indicates operating and maintenance costs. This indicates the cost of purchasing and selling electricity. This indicates the penalty cost for abandoning wind and solar power. Indicates the cost of energy storage losses. This indicates the cost of power supply reliability penalties. This represents the penalty cost for peak-to-valley difference; The constraints of the multi-element energy storage configuration optimization model include power balance constraints, energy storage state of charge constraints, energy storage capacity constraints, energy storage charge and discharge power constraints, thermal energy storage energy balance constraints, hydrogen energy storage quality balance constraints, agricultural machinery mobile energy storage access time constraints, adjustable agricultural load operation time window constraints, and power supply reliability constraints.

9. The method for configuring multi-source energy storage in rural microgrids according to claim 1, characterized in that, In step S7, the annual operating cycle is divided into the off-season, sowing season, irrigation season, harvesting season, drying season or cold chain peak season according to the agricultural production season, and corresponding energy storage operation and scheduling strategies are determined for each season. Among them, during the irrigation season, the available capacity constraint weight of electrochemical energy storage and mobile energy storage of agricultural machinery is increased; during the drying season, the coordinated scheduling weight of thermal energy storage and biomass power generation is increased; during the cold chain peak season, the power supply reliability constraint weight is increased; and during the off-season, the energy storage charge and discharge cycle depth is reduced to reduce energy storage life loss.

10. A multi-element energy storage configuration system for rural microgrids based on quantum-enhanced SARIMA, used to implement the method according to any one of claims 1 to 9, characterized in that, include: The data acquisition module is used to collect rural microgrid operation data, including new energy power generation data, rural domestic load data, agricultural production load data, seasonal agricultural machinery operation data, meteorological data, agricultural production cycle data, and electricity price data. The feature construction module is used to perform time alignment, outlier processing, missing value correction and normalization on the rural microgrid operation data, and to construct a multivariate feature set of the rural microgrid that includes new energy consumption index and seasonal agricultural machinery operation intensity index. The quantum-enhanced prediction module is used to establish a SARIMA prediction model. Based on the multivariate feature set of the rural microgrid, the candidate parameter space and prediction residual of the SARIMA prediction model are corrected. The quantum-enhanced optimization algorithm is used to jointly optimize the non-seasonal and seasonal parameters in the model to obtain the load demand and new energy output prediction results within the target period. The energy storage configuration optimization module is used to establish and solve a multi-element energy storage configuration optimization model based on the load demand and new energy output prediction results, and to determine the capacity, power and operation scheduling strategy of different types of energy storage devices. The solution output module is used to output the multi-energy storage configuration solution for rural microgrids.

Images