A regional micro-grid planning control method and device based on period division

Abstract

The application provides a regional micro-grid planning control method and device based on period division, which comprises the following steps: obtaining power prediction data in a future first planning period of a region, determining micro-grid macro constraint parameters of the first planning period according to the power prediction data; constructing planning benchmark parameters of a second planning period based on the micro-grid macro constraint parameters and the power prediction data; determining operation control parameters of a third planning period based on the planning benchmark parameters; executing the operation control parameters of the third planning period, collecting operation data in execution and generating operation evaluation parameters; correcting the micro-grid macro constraint parameters of the first planning period and / or the planning benchmark parameters of the second planning period according to the operation evaluation parameters; and the planning method can adapt to regional power consumption changes.

Description

Technical Field

[0001] This invention relates to the field of microgrid planning technology, and in particular to a regional microgrid planning and control method and device based on periodic division. Background Technology

[0002] Rural microgrids are core infrastructure supporting rural production and daily life electricity consumption and promoting rural energy transformation. Their planning quality directly affects power supply reliability, renewable energy absorption efficiency, and the effectiveness of low-carbon goals. Currently, rural microgrid planning faces the following prominent problems: Poor load adaptability: The characteristics of large load fluctuations during the busy (irrigation, harvest) and off-seasons in rural areas have not been fully considered. Most of the planning schemes are static designs, which are difficult to dynamically match changes in electricity demand. Insufficient coordination between low-carbon development and supply guarantee: The proportion of new energy sources such as photovoltaics connected to the grid lacks scientific planning, which easily leads to curtailment of solar power. At the same time, there is a conflict between ensuring critical electricity use during the busy farming season and low-carbon goals, and there is a lack of a systematic balancing mechanism. The planning loop is missing: long-term planning, medium-term implementation and short-term scheduling are disconnected, there is no dynamic correction mechanism, and it is difficult to adapt to the changes in electricity consumption structure brought about by rural industrial upgrading and population mobility. Insufficient algorithm adaptability: Traditional planning algorithms are not optimized for scenarios with weak rural data foundations and limited computing resources, resulting in poor practicality and operability.

[0003] Patent CN121146191A discloses a rural microgrid optimization scheduling system and method based on adaptive dynamic programming. Based on real-time data of distributed energy unit power generation and load requirements of load units, a multi-period optimization model of the rural microgrid's state is established using a Markov decision process. An adaptive dynamic programming algorithm, combined with a piecewise linear function approximation function, is used to obtain a multi-period optimization scheduling scheme. This scheme is then used to schedule distributed energy units, energy storage units, and electric vehicle units. While this scheme considers multi-period optimization, it primarily targets the scheduling of distributed energy units, energy storage units, and electric vehicles, and cannot solve the aforementioned problems. Summary of the Invention

[0004] This invention provides a regional microgrid planning and control method and device based on periodic division, which can solve the problems of poor load adaptability, insufficient coordination between low carbon and power supply, lack of closed loop and insufficient algorithm adaptability in existing rural microgrid planning.

[0005] A regional microgrid planning and control method based on periodic partitioning includes: Obtain power forecast data for the region within the first planning period, and determine the macro-constraint parameters of the microgrid for the first planning period based on the power forecast data; Based on the macroscopic constraint parameters of the microgrid and the power forecast data, the planning benchmark parameters for the second planning period are constructed. Based on the aforementioned planning benchmark parameters, the operational control parameters for the third planning cycle are determined; The system executes the operational control parameters for the third planning cycle, collects operational data during execution, and generates operational evaluation parameters. Based on the aforementioned operational evaluation parameters, the macroscopic constraint parameters of the microgrid for the first planning period and / or the planning baseline parameters for the second planning period are revised. The first planning cycle includes the second planning cycle and the first planning cycle is longer than the second planning cycle. The second planning cycle includes the third planning cycle and the second planning cycle is longer than the third planning cycle.

[0006] A regional microgrid planning and control device based on periodic partitioning, comprising: The first planning module is used to acquire power forecast data for the region in the first planning period and determine the macro-constraint parameters of the microgrid in the first planning period based on the power forecast data. The second planning module constructs planning benchmark parameters for the second planning period based on the macro-constraint parameters of the microgrid and the power forecast data. The third planning module determines the operation control parameters for the third planning cycle based on the planning benchmark parameters. The execution module is used to execute the operation control parameters of the third planning cycle, collect the operation data during execution, and generate operation evaluation parameters. The correction module is used to correct the microgrid macro-constraint parameters for the first planning period and / or the planning baseline parameters for the second planning period based on the operational evaluation parameters. The first planning cycle includes the second planning cycle and the first planning cycle is longer than the second planning cycle. The second planning cycle includes the third planning cycle and the second planning cycle is longer than the third planning cycle.

[0007] The regional microgrid planning and control method and apparatus based on periodic partitioning provided by this invention have at least the following beneficial effects: (1) The operation of the data-driven correction mechanism can correct long-term parameters through short-term operation, adapt to changes in electricity consumption brought about by rural industrial upgrading and population mobility in real time, and extend the life cycle of the planning scheme. (2) Strong load adaptability: accurately matches the load fluctuations during the busy and slack farming seasons in rural areas, ensuring key electricity demand; (3) Through phased photovoltaic access planning and dynamic scheduling, the photovoltaic absorption rate is effectively improved, and the annual carbon emission reduction effect meets the requirements of the regional low-carbon policy. (4) The algorithm has been optimized for lightweight design, making it suitable for rural scenarios with limited computing resources. The data requirement is only 2 years of historical load, making it highly operable. Attached Figure Description

[0008] Figure 1 This is a flowchart of one embodiment of the regional microgrid planning and control method based on periodic partitioning provided by the present invention.

[0009] Figure 2 This is a schematic diagram illustrating the photovoltaic absorption rate iteration in an application scenario of the regional microgrid planning and control method based on periodic division provided by the present invention.

[0010] Figure 3 This is a schematic diagram illustrating the photovoltaic access ratio iteration in an application scenario of the regional microgrid planning and control method based on periodic division provided by the present invention.

[0011] Figure 4 This is a schematic diagram illustrating the iterative energy storage capacity under an application scenario of the regional microgrid planning and control method based on periodic partitioning provided by the present invention.

[0012] Figure 5 This is a schematic diagram of the load forecasting interval in an application scenario of the regional microgrid planning and control method based on periodic division provided by the present invention.

[0013] Figure 6 This is a schematic diagram of the cumulative installed capacity in an application scenario of the regional microgrid planning and control method based on periodic division provided by the present invention.

[0014] Figure 7 This diagram illustrates the comparison of operational indicators in an application scenario for the regional microgrid planning and control method based on periodic partitioning provided by the present invention. Detailed Implementation

[0015] To better understand the above technical solutions, the following will provide a detailed explanation of the technical solutions in conjunction with the accompanying drawings and specific implementation methods.

[0016] refer to Figure 1 In some embodiments, a regional microgrid planning and control method based on periodic partitioning is provided, including: S1. Obtain power forecast data for the region in the first planning period, and determine the macro-constraint parameters of the microgrid in the first planning period based on the power forecast data; S2. Based on the macroscopic constraint parameters of the microgrid and the power forecast data, construct the planning benchmark parameters for the second planning period; S3. Based on the planning benchmark parameters, determine the operation control parameters for the third planning cycle; S4. Execute the operation control parameters for the third planning cycle, collect operation data during execution, and generate operation evaluation parameters; S5. Based on the aforementioned operational evaluation parameters, revise the microgrid macro-constraint parameters for the first planning period and / or the planning baseline parameters for the second planning period; The first planning cycle includes the second planning cycle and the first planning cycle is longer than the second planning cycle. The second planning cycle includes the third planning cycle and the second planning cycle is longer than the third planning cycle.

[0017] In some embodiments, the first planning period may be an annual period, the second planning period may be each quarter of that year, and the third planning period may be each month of that year.

[0018] The area in this embodiment can be a rural area.

[0019] Furthermore, in step S1, the power forecast data includes the expected industrial growth intensity, the predicted maximum load value, the predicted photovoltaic power generation, the predicted total power generation of the microgrid, and the predicted energy storage demand for multiple stages within the first planning period of the region.

[0020] Specifically, the projected industrial growth intensity within the first planning cycle can be obtained using trend extrapolation, adapting to scenarios with limited rural industrial data. The predicted maximum load value can be obtained using a grey prediction model. The predicted photovoltaic power generation can be predicted using a simplified BP neural network model. The predicted total microgrid power generation can be predicted using a regression analysis model, combined with historical power generation and load correlation data. Multi-stage predicted energy storage demand can be predicted using a simplified model, correlating load and photovoltaic output fluctuations.

[0021] Based on the power forecast data, the macroscopic constraint parameters of the microgrid for the first planning period are determined, including: S11. Determine the capacity redundancy coefficient based on the projected industry growth intensity; S12. Based on the capacity redundancy coefficient and the predicted maximum load value, calculate the total installed capacity constraint value of the microgrid and the photovoltaic installed capacity constraint value. S13. Calculate the baseline for the photovoltaic grid connection ratio based on the predicted photovoltaic power generation and the predicted total power generation of the microgrid; S14. Calculate the total capacity constraint of the energy storage system based on the predicted energy storage demand at multiple stages.

[0022] Specifically, in step S11, the expected industrial growth intensity can be the expected industrial output growth rate of the region.

[0023] Determine the reasonable range of capacity redundancy coefficient [k] based on industry standards for microgrid planning. min k max ], where kmin k is the lower limit of the capacity redundancy coefficient (adapting to low-growth scenarios and ensuring basic power supply redundancy). max This is the upper limit of the capacity redundancy coefficient (to adapt to high-growth scenarios and cope with surges in load demand).

[0024] Furthermore, the capacity redundancy coefficient can be determined using linear interpolation: (1) Where, k red r is the capacity redundancy factor. min r is the benchmark value for the annual industrial output growth rate. max Let r be the critical value for the annual industry output growth rate, and r be the current projected annual industry output growth rate. If r ≥ r max Then take r=r max If r≤r min Then take r=r min .

[0025] In step S12, the total installed capacity constraint value of the microgrid is calculated using the following formula: (2) Among them, P total-lim k represents the total installed capacity constraint value of the microgrid. red P represents the capacity redundancy factor. long-max This represents the predicted maximum load value within the first planning period.

[0026] Taking a one-year planning cycle as an example: (3) Among them, P long-max(year) P represents the projected maximum load value for the first planning period (one year). q1-max P represents the projected maximum load value for the first quarter. q2-max P represents the projected maximum load value for the second quarter. q2-max P represents the projected maximum load value for the third quarter. q4-max This indicates the projected maximum load value for the fourth quarter.

[0027] Furthermore, based on the total installed capacity constraint of the microgrid and the proportion of photovoltaic installations in the region, the photovoltaic installed capacity constraint can be calculated.

[0028] Furthermore, the aforementioned predicted maximum load value is obtained based on a pre-built grey prediction model, which predicts the predicted maximum load value by including: Historical load data for n years within the collection area L={l1,l2,……l n} (n≥2, including load data for busy and off-season farming), li The historical load data for year i is obtained directly from the rural statistics department. The historical load data L for the past n years should include quarterly peak / valley load data (in kW) for typical busy and slack farming seasons. The data collection frequency should not be lower than the minimum frequency required by the rural electricity consumption statistics standards, and the data should be obtained directly from the electricity consumption statistics ledger of the rural power supply station.

[0029] Input the historical load data of n years in the region into the grey prediction model; The grey prediction model performs an accumulation operation based on the historical load data to obtain the accumulated sequence: (4) Construct matrix B and vector Y based on the accumulated sequence: (5) (6) Based on the given matrix and vector, the parameter vector is solved using the least squares method. : (7) Where a is the development coefficient and b is the gray action amount, obtained through calculation.

[0030] Based on the parameter vector and the accumulated sequence, prediction and inverse accumulation operations are performed to obtain the predicted maximum load value: (8) (9) (10) In step S13, the baseline for the photovoltaic grid connection ratio is the ratio of the predicted photovoltaic power generation to the predicted total power generation of the microgrid.

[0031] In step S14, the total capacity constraint of the energy storage system is: (11) in, To meet the annual quarterly energy storage demand, the quarterly electricity gap in the second planning cycle (medium-term planning) can be used to deduce that a set of energy storage equipment can dynamically adapt to the annual load fluctuations.

[0032] Total capacity constraints of energy storage systems The maximum energy storage configuration during the second planning cycle (quarterly for medium-term planning) must not exceed [a certain value]. The determined photovoltaic installed capacity must meet the requirement that the total photovoltaic output targets for the second planning cycle (quarterly in the medium-term plan) do not exceed the maximum output of the photovoltaic installed capacity, and ensure the feasibility of photovoltaic power consumption projects.

[0033] Further, in step S2, based on the microgrid macro-constraint parameters of the first planning period and the power forecast data, the planning benchmark parameters for the second planning period are constructed, including: S21. Calculate the historical load factor based on the historical load data of the region; S22. Based on the historical load factor and the predicted maximum load value, construct the load forecast interval for the second planning period; S23. Based on the photovoltaic access ratio baseline and the total electricity demand of the microgrid in the second planning period, calculate the photovoltaic output target for the second planning period. S24. Determine the optimal range for energy storage configuration based on the total capacity constraint of the energy storage system; S25. Generate a priority list for ensuring power supply based on regional electricity demand; S26. Based on the energy storage configuration optimization interval and the photovoltaic access ratio baseline, a non-dominated sorting genetic algorithm is used for multi-objective optimization to obtain the optimal photovoltaic access ratio and optimal energy storage configuration within the second planning period.

[0034] Specifically, in step S21, the historical load factor includes the ratio of the load trough value of the same historical period to the maximum load value of the whole year (minimum load factor), and the ratio of the load peak value of the same historical period to the maximum load value of the whole year (maximum load factor).

[0035] In step S22, the load forecast interval for the second planning period (quarter) [P] q-min P q-max The formula for calculating ] is: (12) (13) in, The predicted maximum load value determined in step S1 (i.e., the maximum value of the upper limit of the quarterly load range for that year). The minimum load factor for the quarter (for the off-season, the load trough value of the same quarter of the previous year and the maximum load value of the microgrid in the previous year are taken as the minimum load factor). (ratio) The quarterly maximum load factor (for the busy farming season, the peak load of the same quarter of the previous year and the maximum load value of the microgrid in the previous year are taken as the factor). The ratios (both based on load data from the same quarter of the previous year) are all derived from the load data of the same quarter of the previous year. Compared with the previous year's maximum microgrid load value The ratio is calculated as follows, i.e. (14) (15) The "peak / valley load" refers to the maximum / minimum power consumption per unit time, in kW; the previous year's load data and the previous year's maximum microgrid load value are directly obtained from the annual operation report of the microgrid operation monitoring system.

[0036] In step S23, the photovoltaic output target Q for the second planning period (by quarter) q-target The calculation formula is: (16) in, The total quarterly electricity demand of the microgrid (in kWh) is calculated by multiplying the average value Pq-avg of the quarterly load forecast interval by the quarterly electricity consumption duration Tq: (17) (18) Quarterly electricity usage time The duration of the busy farming season is determined based on the intensity of agricultural production, while the duration of the slack farming season is determined based on the patterns of residential electricity consumption. The data is directly obtained from the production and living statistics of the village committees.

[0037] busy farming season The target is increased by a preset ratio based on the above formula, and the quarterly photovoltaic output target must be compatible with the photovoltaic installed capacity constraint value in step S13.

[0038] In step S24, based on the total capacity constraint S of the energy storage system... long-max The optimal range for energy storage configuration is determined to be [0, S]. long-max Furthermore, it must meet the requirement that 'the proportion of energy storage configuration during the busy farming season is greater than or equal to the preset proportion of total annual energy storage' (to meet the high load demand during the busy farming season).

[0039] Furthermore, in step S25, the priority list for supply is determined based on the core needs of rural production and life, and the general order is "basic electricity consumption for rural residents (including basic electricity consumption for lighting, cooking and other basic needs) > key agricultural electricity consumption during the busy farming season (including core links such as irrigation, sowing and harvesting) > commercial electricity consumption during the off-season (including small shops, non-core processing and other aspects)". Information on key agricultural electricity consumption links is directly obtained from the annual work plan of the rural agricultural department.

[0040] In step S6, based on the baseline of the photovoltaic grid connection ratio, the optimal range for the photovoltaic grid connection ratio is determined as follows: ,in, The allowable fluctuation range is determined based on the regional low-carbon plan.

[0041] The multi-objective optimization was determined using the non-dominated sorting genetic algorithm NSGA-II, with the optimization variables being: That is, the proportion of photovoltaic grid connection in each second planning cycle (quarter): And the energy storage configuration for each second planning cycle (quarter): The optimization targets include the number of supply gaps during the busy farming season, photovoltaic grid integration rate, and economic costs.

[0042] Furthermore, a non-dominated sorting genetic algorithm is used for multi-objective optimization, including: Initialize the population, where each individual in the population is a combination of energy storage configuration and photovoltaic access ratio; The population is sorted into non-dominated layers to obtain multiple non-dominated layers; Calculate the crowding level of individuals in each non-dominated layer; The selection probability is calculated based on the crowding level and non-dominant rank of individuals, and the parent population is selected based on the selection probability. Based on the parent population, crossover and mutation operations are performed to obtain the offspring population; The parent and offspring populations are merged to obtain a new generation population, which is then iterated until the stopping condition is met.

[0043] Specifically, an initial population of size N is established (N is a small value that satisfies the computational efficiency of rural scenarios, determined based on the computing power of ordinary office equipment, and directly obtained from the microgrid planning algorithm application guide), and each individual in the population is a combination of decision variables. ; in, The proportion of photovoltaic grid connection for each second planning cycle (quarter), For each second planning cycle (quarter), the energy storage configuration shall be such that the energy storage configuration capacity during the busy farming season accounts for no less than the preset proportion of the total annual energy storage, and the maximum value of the energy storage configuration capacity for all quarters does not exceed [a certain percentage]. ); The population is subjected to non-dominated ranking, and different non-dominated layers are formed. The non-dominated ranking is based on the following: if all the optimization objective function values ​​of individual A are... All meet And there exists at least one m such that If individual A dominates individual B; the objective function for optimization includes "the number of supply gaps during the busy farming season ≤ the allowable threshold" (the allowable threshold is directly obtained from the rural agricultural production electricity guarantee standard). Calculate the crowding level of individuals in each non-dominated layer. The formula is: (19) in, Let be the function value of individual i on the m-th objective. , The maximum and minimum values ​​of the m-th target in the current undominated layer are obtained through calculation; The tournament selection method is used to select outstanding individuals. The selection probability is positively correlated with the non-dominated layer ranking and the crowding degree. The higher the non-dominated layer ranking and the greater the crowding degree, the higher the selection probability. The selected individuals undergo simulated binary crossover (the crossover probability is taken as a standard value suitable for a small population, directly obtained from the basic application specifications of genetic algorithms). The crossover formula is: If random number ,but (20) otherwise: ;(twenty one) Simultaneously, a polynomial mutation operation is performed (the mutation probability is taken as a conventional value suitable for a small population, directly obtained from the basic application specifications of genetic algorithms). The mutation formula is: If random number

[0044] ;(twenty three) otherwise: ;(twenty four) in, For parental individual decision variables, For the decision-making variables of offspring individuals, For the decision variables of the individual to be mutated, (These are the individual decision variables after the mutation). The parent and offspring populations are merged to a size of 2N. The top N optimal individuals are selected through non-dominated sorting and crowding calculation to form a new generation population. This process is repeated until the maximum number of iterations is reached (the maximum number of iterations is a small value that meets the convergence accuracy requirements, determined according to the computational efficiency requirements of the rural scenario) or the change in the objective function value of the optimal solution between two adjacent generations is less than the convergence threshold (the convergence threshold is directly obtained from the accuracy standard of the microgrid planning algorithm). The optimal combination of decision variables is then output.

[0045] The population size N of the non-dominated sorting genetic algorithm NSGA-II is set to 20-30, the maximum number of iterations is set to 50-60, and the crossover and mutation probabilities are taken as conventional values ​​suitable for this population size, all directly obtained from the Rural Microgrid Lightweight Algorithm Application Manual; and the algorithm optimization objective must include "the total capacity constraint s of the energy storage system in step S1, with the quarterly energy storage configuration not exceeding the total capacity of the energy storage system in step S1". long-max "and the photovoltaic installed capacity constraint value of the quarterly photovoltaic output target adaptation step S13" to ensure the consistency between medium-term planning parameters and long-term planning macro constraints.

[0046] Further, in step S3, based on the planning benchmark parameters, the operation control parameters for the third planning cycle are determined, including: S31. Based on the load forecast interval, calculate the equipment operating power for the third planning period; S32. Construct a discrete-time state-space model of the microgrid, and use the optimal energy storage configuration and photovoltaic output target of the second planning period as constraints of the discrete-time state-space model of the microgrid. S33. Based on the discrete-time state-space model of the microgrid, perform state prediction and solve for the operation control vector sequence within the third planning period; S34. Determine the photovoltaic dispatch strategy based on the aforementioned supply priority list.

[0047] Specifically, in step S31, the equipment operating power for the third planning period (monthly) is based on the load forecast for the third planning period (monthly). (Unit: kW) The load forecast value for the third planning period is determined. The calculation formula is: (25) In the formula: (26) P q-ref The average load forecast range (in kW) for the second planning period (quarter) is obtained through calculation. This is the quarterly load deviation correction factor, calculated using the average actual load of the previous quarter. (Unit: kW) and the average load forecast of the previous quarter The deviation rate (in kW) is calculated as follows: (27) busy farming season Based on the above formula, an additional agricultural situation correction amount is added (the correction amount ratio is determined according to the electricity consumption fluctuation characteristics of the busy farming month, and is directly obtained from the monthly electricity consumption statistics report of the village; the average actual load of the previous quarter and the average load forecast of the previous quarter are directly obtained from the quarterly operation report of the microgrid operation monitoring system).

[0048] Furthermore, in step S32, the microgrid discrete-time state-space model includes a state vector comprising a load correction term, a state of charge of energy storage, and a predicted photovoltaic output value; a control vector comprising energy storage charging power, energy storage discharging power, and grid power purchase; a disturbance vector comprising load disturbance and photovoltaic output disturbance; and an output vector comprising a predicted power supply load and a predicted photovoltaic absorption rate.

[0049] In step S33, the system state at future times is predicted based on the current state vector and the discrete-time state-space model of the microgrid; and the control vector sequence that minimizes the objective function is solved based on the system state at future times.

[0050] Specifically, the discrete-time state-space model of a microgrid is shown below: (28) in, (29) x(k) is the state vector, L pred (k) is the load correction term (including busy farming months). In the state of energy storage charge, Forecast value of photovoltaic power output; (30) u(k) is the control vector. For energy storage charging power, For energy storage discharge power, Power purchased from the power grid; (31) Let be the perturbation vector. For load disturbance, Disturbance to photovoltaic output; (32) For the output vector, To predict power supply load, To predict the photovoltaic absorption rate; A, B, C, and D are system matrices, determined through system identification or microgrid operation experience.

[0051] Based on the current state x(k) and the state-space model, predict the future. The system status is determined by taking a small-scale time-domain value that is suitable for short-term adjustments, based on the monthly electricity stability requirements of rural areas. (33) Construct the objective function: (34) In the formula: Weighting coefficients ( (A larger value is taken during the busy farming season), determined based on the needs of supply guarantee, low carbon, and economic goals; For the control time domain (taking a small-scale value adapted to the prediction time domain); R is the weight matrix; in the prediction time domain, the sequence of control vectors that minimizes the objective function J is solved using an optimization algorithm (such as quadratic programming): (35) Output the optimal control strategy u(k) at the current moment.

[0052] Furthermore, in step S34, the scheduling priority of the photovoltaic dispatch strategy is based on the photovoltaic absorption rate of the previous second planning cycle (previous quarter). Adjustments to the photovoltaic grid integration rate mentioned in the second planning cycle (previous quarter). The calculation formula is: (35) Among them, Q q-act This refers to the actual photovoltaic power consumed in the previous second planning cycle (the previous quarter). The figures represent the actual photovoltaic power generation in the previous second planning cycle (last quarter), all in kWh, directly obtained from the microgrid metering system. Furthermore, photovoltaic dispatching needs to be adapted to the photovoltaic installed capacity of the first planning cycle and the photovoltaic output target of the second planning cycle.

[0053] The universal dispatch logic is as follows: "Photovoltaic power output is prioritized to supply the highest priority load, and the remaining output is used for energy storage charging; energy storage discharges during peak load periods, prioritizing high priority loads in the supply guarantee list; grid power purchase serves as a final supplement and is only activated when photovoltaic and energy storage outputs are insufficient"; and the energy storage discharge power must meet the total energy storage capacity constraint of the long-term planning in step S1. This ensures that the state of charge (SOC) of the energy storage is within a safe range.

[0054] Further, in step S4, runtime data is collected and runtime evaluation parameters are generated, including: S41. Obtain the actual average load within multiple third planning periods, and calculate the actual load deviation rate based on the actual average load and the load prediction interval within the second planning period. S42. Obtain microgrid fault records and count the number of supply gaps and power shortages within multiple third planning cycles; S43. Statistically analyze the actual photovoltaic power consumption and actual photovoltaic power generation in multiple third planning cycles, and calculate the actual photovoltaic consumption rate; S44. Calculate the total energy storage discharge within multiple third planning cycles and the actual energy storage utilization rate.

[0055] Specifically, in step S41, the quarterly actual load deviation rate The formula used to verify the accuracy of load forecasting is: (36) In the formula: P q-act The actual average load (in kW, taken from the microgrid operation monitoring system) for multiple third planning cycles (quarters). The average load forecast for the corresponding quarter in the second planning period: ;(37)

[0056] In step S42, the number of times the power supply for "basic residential electricity consumption / critical agricultural electricity consumption during busy farming seasons" was insufficient within multiple third planning cycles is counted (taken from microgrid fault records).

[0057] In step S43, the actual photovoltaic grid integration rate within multiple third planning cycles (quarters) The formula used to test photovoltaic utilization efficiency is: (38) In the formula, Actual photovoltaic power consumption (unit: kWh, taken from microgrid metering system) for multiple third planning cycles (quarters). The actual photovoltaic power generation (in kWh, taken from the microgrid metering system) for multiple third planning cycles (quarters).

[0058] In step S44, the actual energy storage capacity utilization rate The formula used to test the efficiency of energy storage resources is as follows: (39) In the formula, Total energy storage discharge (in kWh) over multiple third planning cycles (quarters). The total capacity constraint of the energy storage system determined for the first planning cycle (unit: kWh), 90 represents the average number of days in a quarter (simplified calculation).

[0059] Further, in step S5, based on the operational evaluation parameters, the macroscopic constraint parameters of the microgrid for the first planning period and / or the planning baseline parameters for the second planning period are corrected, including: S51. If the average of the sum of the number of supply gaps and the number of power shortages in multiple third planning cycles is greater than the first preset value, then the capacity redundancy coefficient of the next first planning cycle shall be adjusted. S52. If the actual load deviation rate is greater than the first preset value, then the load forecast range for the next second planning period shall be corrected according to the actual load deviation rate. S53. If the actual photovoltaic absorption rate is less than the second preset value, then the photovoltaic output target for the next second planning cycle shall be adjusted according to the photovoltaic absorption rate. S54. If the actual utilization rate of the energy storage is lower than the third preset value, the total installed capacity constraint value of the microgrid in the next first planning cycle shall be adjusted according to the actual utilization rate of the energy storage. S55, calculate the average photovoltaic absorption rate and the average actual energy storage utilization rate within at least two second planning periods. If the average photovoltaic absorption rate is less than the second preset value, or the average actual energy storage utilization rate is less than the third preset value, then simultaneously adjust the photovoltaic access ratio baseline and the total capacity constraint of the energy storage system for the next first planning period.

[0060] Specifically, in step S51, if the average sum of the number of supply gaps and the number of power shortages in multiple third planning cycles is greater than the first preset value, the capacity redundancy coefficient is reduced proportionally, and the reduced capacity redundancy coefficient is not lower than the safety limit of the rural microgrid capacity, all of which are directly obtained from the microgrid parameter adjustment specifications.

[0061] In step S52, the calculation formula for the load forecast interval of the second planning period is revised as follows: (40) Among them, P q-min ' is the minimum value of the corrected load forecast interval, P q-max ' is the maximum value of the corrected load forecast interval, P q-min P is the minimum value of the load forecast interval before correction. q-max This represents the maximum value of the load forecast range before correction. And P q-max 'Less than or equal to the predicted maximum load value of the next first planning cycle.'

[0062] In step S53, the photovoltaic power output target for the next second planning cycle is revised, and the calculation formula is as follows: (41) Among them, Q q-target The revised photovoltaic power output target is less than or equal to the maximum photovoltaic installed capacity output in the next first planning cycle. This is the third preset value.

[0063] In step S54, the total capacity constraint of the energy storage system in the next first planning cycle is corrected, and the calculation formula is as follows: (42) Among them, S long-max 'This is the revised total capacity constraint for the energy storage system.' This is the fourth preset value.

[0064] In step S55, the baseline for the photovoltaic grid connection ratio and the total capacity constraint of the energy storage system for the next first planning cycle are revised, and the revision formulas are as follows: (43) (44) Where, α base'This is the revised baseline for the proportion of photovoltaic power connected to the grid.' Average photovoltaic grid integration rate This represents the average actual utilization rate of energy storage.

[0065] The methods provided in the above embodiments will be further explained through specific application scenarios below.

[0066] Phased planning of a rural low-carbon microgrid: Historical data: Quarterly energy consumption data for 2023 and 2024 (actual measurements by the power supply station, including load characteristics during busy / slack farming seasons); Planning years: Q1 2025 (short-term), 2025 (medium-term), 2025-2027 (long-term). All planning data are forecasts based on historical data.

[0067] Key measured data: Maximum load L = 260200280230 kW in Q1-Q4 2023; Maximum load L = 270210300250 kW in Q1-Q4 2023 (meets the patent requirement of "n≥2 years"); Peak and valley loads for the first four quarters of 2024: Q1=300kW / 250kW, Q2=240kW / 200kW, Q3=320kW / 280kW, Q4=270kW / 230kW (taken from the "2024 Microgrid Operation Annual Report"). The maximum load in 2024 is Lmax-last = 320kW.

[0068] Photovoltaic resources: 1150 hours of effective utilization per year (measured in 2023-2024, used for quarterly photovoltaic output calculation); energy storage is subject to the general rule of operation, which stipulates a maximum of two charging and two discharging cycles per day.

[0069] Policy and standard parameters: Capacity redundancy factor k red =1.2, Baseline α for photovoltaic grid connection ratio base =40%, permissible threshold for load deviation Photovoltaic grid integration threshold N times allowed to fill the supply gap during busy farming season q-lim =1 time / quarter; Economic parameters: Photovoltaic power generation cost: 0.35 yuan / kWh; Grid electricity purchase cost: 0.52 yuan / kWh; Energy storage operation and maintenance cost: 1.0 yuan / kWh. (Market price in 2024).

[0070] Long-term planning calculations: The grey forecasting model calculates the upper limit P of the quarterly load forecasting interval. q-max The historical load is calculated by accumulating the historical load data for the eight quarters of 2023-2024: L=[260,200,280,230,270,210,300,250]; Perform one accumulation to generate the accumulation sequence: ; Final cumulative sequence: L(1)=[260,460,740,970,1240,1450,1750,2000]; Formula for solving parameters: , where a is the development coefficient and b is the gray action amount.

[0071] ; The final solution yields the following parameters: development coefficient a ≈ -0.022, gray action amount b ≈ 246. Forecasted quarterly load range upper limit The formula for predicting cumulative sequences is: ; Where L(1) = 260 (initial load value), a = -0.022, b = 246, and k is the prediction step size (3 years from 2025 to 2027, 1 step size per quarter, so k = 9 to k = 20, corresponding to 2025Q1 to 2027Q4, that is, from the first quarter of 2025 to the fourth quarter of 2027).

[0072] (Predict and reconstruct the quarterly cumulative sequence for 2025-2027 using the inverse cumulative operation:) ; Restoring quarterly load (i.e.) ).

[0073] Calculate the maximum annual load: 2025 (k=4): P long-max-2025 =l (1) (5) -l (1) (4)≈305kW; 2026 (k=8): P long-max-2026 =l (1) (9)-l (1) (8)≈335kW; 2027 (k=12): P long-max-2027 =l (1) (13) -l (1) (12)≈370kW.

[0074] Photovoltaic installations will account for 30% of total installed capacity. Photovoltaic installed capacity projections for 2025, 2026, and 2027 are as follows: P pv-toatl-2025 =kred*Plong-max-2025 *30%=1.2*305*30%≈110kW; P pv-toatl-2026 =kred*P long-max-2026 *30%=1.2*335*30%≈120kW; P pv-toatl-2027 =kred*P long-max-2026 *30%=1.2*370*30%≈133kW.

[0075] Medium-term planning calculations: Calculate the quarterly load forecast interval [P] based on formulas (12) to (15) and the data in Table 1. q-min P q-max ], Table 1

[0076] The quarterly load forecast intervals are shown in Table 2: Table 2

[0077] Based on the total photovoltaic (PV) installed capacity of 366kW in 2025, the cumulative quarterly PV installed capacity (cumulative quarterly, with subsequent quarters ≥ previous quarters, and Q1 cumulative capacity being the upper limit of that quarter's installed capacity) is shown in Table 3: Table 3

[0078] The quarterly photovoltaic power generation (matching 1150h, cumulative installed capacity × quarterly utilization hours) is shown in Table 4: Table 4

[0079] The photovoltaic power output targets are shown in Table 5: ; in, This refers to the duration of electricity consumption per quarter.

[0080] Table 5

[0081] Autonomous optimization using the NSGA-II algorithm Decision variable α q , (Accumulated energy storage installations are optimized by the algorithm, with subsequent quarters ≥ previous quarters, and the cumulative installations in Q1 being the upper limit of that quarter).

[0082] Cost function: ; Crowding calculation:

[0083] The complete 30-step iteration table of the NSGA-II algorithm is shown in Table 6 (taking Q1 as an example): Table 6

[0084] The iterative curves of photovoltaic absorption rate, photovoltaic grid connection ratio and energy storage installed capacity are as follows: Figure 2 , Figure 3 , Figure 4 As shown.

[0085] Similarly, the optimization for Q2-Q4 was completed, and the cumulative installed capacity and total capacity of energy storage in 2025 were finally derived (Q1 cumulative capacity is the upper limit of the quarter's installed capacity), as shown in Table 7: Table 7

[0086] The calculation results for the medium-term plan are shown in Table 8 (quarterly level, cumulative installed capacity in 2025): Table 8

[0087] Medium-term planning load range, cumulative photovoltaic installed capacity, such as Figure 5 and Figure 6 As shown.

[0088] Short-term planning: Monthly load forecast: Quarterly load reference values: ; Deviation correction factor: ; P q-act-last =250kW; (Actual measurement in Q1 2024) Monthly load: ;

[0089] Table 9 shows the monthly cumulative photovoltaic installed capacity and power generation (based on a cumulative photovoltaic installed capacity of 90kW in mid-Q1 2025, with monthly cumulative ≤ quarterly cumulative): Table 9

[0090] Table 10 shows the monthly cumulative installed capacity and charge / discharge of energy storage (based on a cumulative installed capacity of 50kWh in mid-Q1 2025, with monthly cumulative ≤ quarterly cumulative). Table 10

[0091] Quadratic programming algorithm (MPC objective function optimization): Objective function: ; The directly obtained data are: w1=0.6, w2=0.3, w3=0.1; u=[P Es-ch P ES-dis P grid ] T (Control Vector) Constraint power balance: E pv +P ES-ch -P Es-dis =L pred +P grid ; Energy storage SOC ∈ [30%, 100%] The complete 30-step iteration of the quadratic programming is shown in Table 11 (taking January 2025 as an example). Table 11

[0092] Similarly, the February and March iterations were completed, and the final short-term planning results are shown in Table 12 below (monthly level for Q1 2025, cumulative installations, monthly cumulative ≤ quarterly cumulative): Table 12

[0093] Short-term operational testing and evaluation (after Q1 2025): Measured data was collected through the microgrid intelligent monitoring system, and evaluation indicators were calculated using the patented formula. Conclusions were drawn after comparing these indicators with threshold values. Table 13

[0094] For example, each indicator Figure 7 As shown, all indicators meet the standards, and there is no need to adjust the planning parameters.

[0095] The next quarter (Q2 2025) will continue to use the cumulative installed capacity parameters of the medium-term plan: load range [191,229] kW, cumulative photovoltaic installed capacity of 180 kW, and cumulative energy storage installed capacity of 70 kWh.

[0096] Photovoltaic and energy storage installations will continue to be accumulated on a quarterly / monthly basis (monthly cumulative ≤ quarterly cumulative, quarterly cumulative ≤ annual total installations), and the installation volume will only increase in subsequent stages; if subsequent indicators fail to meet the standards, corrections will be made according to step S5.

[0097] In some embodiments, a regional microgrid planning and control device based on periodic partitioning is also provided, comprising: The first planning module is used to acquire power forecast data for the region in the first planning period and determine the macro-constraint parameters of the microgrid in the first planning period based on the power forecast data. The second planning module constructs planning benchmark parameters for the second planning period based on the macro-constraint parameters of the microgrid and the power forecast data. The third planning module determines the operation control parameters for the third planning cycle based on the planning benchmark parameters. The execution module is used to execute the operation control parameters of the third planning cycle, collect the operation data during execution, and generate operation evaluation parameters. The correction module is used to correct the microgrid macro-constraint parameters for the first planning period and / or the planning baseline parameters for the second planning period based on the operational evaluation parameters. The first planning cycle includes the second planning cycle and the first planning cycle is longer than the second planning cycle. The second planning cycle includes the third planning cycle and the second planning cycle is longer than the third planning cycle.

[0098] Please refer to the method implementation examples for the functions of each module; they will not be repeated here.

[0099] The regional microgrid planning and control method and apparatus based on periodic partitioning provided in the above embodiments have at least the following beneficial effects: (1) The operation of the data-driven correction mechanism can correct long-term parameters through short-term operation, adapt to changes in electricity consumption brought about by rural industrial upgrading and population mobility in real time, and extend the life cycle of the planning scheme. (2) Strong load adaptability: accurately matches the load fluctuations during the busy and slack farming seasons in rural areas, ensuring key electricity demand; (3) Through phased photovoltaic access planning and dynamic scheduling, the photovoltaic absorption rate is effectively improved, and the annual carbon emission reduction effect meets the requirements of the regional low-carbon policy. (4) The algorithm has been optimized for lightweight design, making it suitable for rural scenarios with limited computing resources. The data requirement is only 2 years of historical load, making it highly operable.

[0100] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention. Clearly, those skilled in the art can make various alterations and modifications to the invention without departing from its spirit and scope. Thus, if these modifications and modifications of the invention fall within the scope of the claims and their equivalents, the invention is also intended to include these modifications and modifications.

Claims

1. A regional microgrid planning and control method based on periodic partitioning, characterized in that, include: Obtain power forecast data for the region within the first planning period, and determine the macro-constraint parameters of the microgrid for the first planning period based on the power forecast data; Based on the macroscopic constraint parameters of the microgrid and the power forecast data, the planning benchmark parameters for the second planning period are constructed. Based on the aforementioned planning benchmark parameters, the operational control parameters for the third planning cycle are determined; The system executes the operational control parameters for the third planning cycle, collects operational data during execution, and generates operational evaluation parameters. Based on the aforementioned operational evaluation parameters, the macroscopic constraint parameters of the microgrid for the first planning period and / or the planning baseline parameters for the second planning period are revised. The first planning cycle includes the second planning cycle and the first planning cycle is longer than the second planning cycle. The second planning cycle includes the third planning cycle and the second planning cycle is longer than the third planning cycle.

2. The method according to claim 1, characterized in that, The power forecast data includes the projected industrial growth intensity, projected maximum load value, projected photovoltaic power generation, projected total microgrid power generation, and multi-stage projected energy storage demand during the first planning cycle within the region. Based on the power forecast data, the macroscopic constraint parameters of the microgrid for the first planning period are determined, including: The capacity redundancy coefficient is determined based on the projected industry growth intensity. Based on the capacity redundancy coefficient and the predicted maximum load value, the total installed capacity constraint value of the microgrid and the photovoltaic installed capacity constraint value are calculated. The baseline for the photovoltaic grid connection ratio is calculated based on the predicted photovoltaic power generation and the predicted total power generation of the microgrid. The total capacity constraint of the energy storage system is calculated based on the predicted energy storage demand at multiple stages.

3. The method according to claim 2, characterized in that, The projected industrial growth intensity is the projected industrial output growth rate within the region; the capacity redundancy coefficient is calculated using linear interpolation based on the projected industrial output growth rate. The total installed capacity constraint value of the microgrid is the product of the capacity redundancy coefficient and the predicted maximum load value; The photovoltaic grid connection ratio baseline is the ratio of predicted photovoltaic power generation to predicted total microgrid power generation; The total capacity constraint of the energy storage system is the maximum value among the predicted energy storage demands in multiple stages.

4. The method according to claim 2, characterized in that, Based on the macroscopic constraint parameters of the microgrid and the power forecast data, planning baseline parameters for the second planning period are constructed, including: Calculate the historical load factor based on the historical load data of the region; Based on the historical load factor and the predicted maximum load value, a load forecast interval for the second planning period is constructed; Based on the aforementioned photovoltaic grid connection ratio baseline and the total electricity demand of the microgrid in the second planning period, the photovoltaic output target for the second planning period is calculated. Based on the total capacity constraint of the energy storage system, the optimal range for energy storage configuration is determined; A priority list for ensuring power supply is generated based on regional electricity demand. Based on the energy storage configuration optimization interval and the photovoltaic access ratio baseline, a non-dominated sorting genetic algorithm is used for multi-objective optimization to obtain the optimal photovoltaic access ratio and optimal energy storage configuration within the second planning period.

5. The method according to claim 4, characterized in that, Multi-objective optimization is performed using a non-dominated sorting genetic algorithm, including: Initialize the population, where each individual in the population is a combination of energy storage configuration and photovoltaic access ratio; The population is sorted into non-dominated layers to obtain multiple non-dominated layers; Calculate the crowding level of individuals in each non-dominated layer; The selection probability is calculated based on the crowding level and non-dominant rank of individuals, and the parent population is selected based on the selection probability. Based on the parent population, crossover and mutation operations are performed to obtain the offspring population; The parent and offspring populations are merged to obtain a new generation population, which is then iterated until the stopping condition is met.

6. The method according to claim 4, characterized in that, Based on the planning baseline parameters of the second planning period, the operational control parameters for the third planning period are determined, including: Based on the load forecast interval, the equipment operating power for the third planning period is calculated. A discrete-time state-space model of a microgrid is constructed, with the optimal energy storage configuration and photovoltaic output target for the second planning period as constraints on the discrete-time state-space model of the microgrid. State prediction is performed based on the microgrid discrete-time state-space model, and the operation control vector sequence within the third planning period is solved. The photovoltaic dispatch strategy is determined based on the aforementioned supply priority list.

7. The method according to claim 4, characterized in that, Collect runtime data and generate runtime evaluation parameters, including: Obtain the actual average load within multiple third planning periods, and calculate the actual load deviation rate based on the actual average load and the load prediction interval within the second planning period; Acquire microgrid fault records and count the number of supply gaps and power shortages within multiple third planning cycles; The actual photovoltaic power consumption and actual photovoltaic power generation during multiple third planning cycles are statistically analyzed to calculate the actual photovoltaic power consumption rate. The total discharge of energy storage during multiple third planning cycles is statistically analyzed to calculate the actual utilization rate of energy storage.

8. The method according to claim 7, characterized in that, Based on the aforementioned operational evaluation parameters, the macroscopic constraint parameters of the microgrid for the first planning period and / or the planning baseline parameters for the second planning period are revised, including: If the average of the sum of the number of supply gaps and the number of power shortages in multiple third planning cycles is greater than the first preset value, then the capacity redundancy coefficient of the next first planning cycle will be adjusted. If the actual load deviation rate is greater than the second preset value, the load forecast range for the next second planning period is adjusted according to the actual load deviation rate. If the actual photovoltaic absorption rate is less than the third preset value, the photovoltaic output target for the next second planning cycle will be adjusted based on the photovoltaic absorption rate. If the actual utilization rate of energy storage is lower than the fourth preset value, the total installed capacity constraint value of the microgrid for the next first planning cycle shall be adjusted according to the actual utilization rate of energy storage.

9. The method according to claim 8, characterized in that, After calculating the actual photovoltaic absorption rate and the actual energy storage utilization rate, the following is also included: Calculate the average photovoltaic absorption rate and the average actual energy storage utilization rate over at least two second planning periods. If the average photovoltaic absorption rate is less than the third preset value, or the average actual energy storage utilization rate is less than the fourth preset value, then the baseline of photovoltaic access ratio and the total capacity constraint of energy storage system for the next first planning period are simultaneously adjusted.

10. A regional microgrid planning and control device based on periodic partitioning, characterized in that, include: The first planning module is used to acquire power forecast data for the region in the first planning period and determine the macro-constraint parameters of the microgrid in the first planning period based on the power forecast data. The second planning module constructs planning benchmark parameters for the second planning period based on the macro-constraint parameters of the microgrid and the power forecast data. The third planning module determines the operation control parameters for the third planning cycle based on the planning benchmark parameters. The execution module is used to execute the operation control parameters of the third planning cycle, collect the operation data during execution, and generate operation evaluation parameters. The correction module is used to correct the microgrid macro-constraint parameters for the first planning period and / or the planning baseline parameters for the second planning period based on the operational evaluation parameters. The first planning cycle includes the second planning cycle and the first planning cycle is longer than the second planning cycle. The second planning cycle includes the third planning cycle and the second planning cycle is longer than the third planning cycle.

Images