Island micro-grid optimal dispatching and oil supplement decision method and system

Abstract

The application discloses an island micro-grid optimal scheduling and oil supplement decision method and system. Based on state information at a current scheduling moment, load prediction and renewable energy output prediction in a future time window are determined, a net load gap, energy supply ability replaceable by energy storage and minimum diesel demand are determined, a dynamic safety inventory boundary is constructed, and a mixed integer linear programming joint optimization model considering diesel engine start-stop, diesel engine output, energy storage charging and discharging, general load reduction and oil supplement action is established, so that optimal scheduling parameters are directly output to perform corresponding control on a micro-grid system. The application can realize joint optimization of energy storage regulation, diesel engine scheduling, load management and oil supplement action of the island micro-grid, and realize collaborative optimization of the energy storage and the diesel engine.

Description

Technical Field

[0001] This application relates to the field of microgrid operation control and optimal scheduling technology, and in particular to a method and system for optimal scheduling and refueling decision-making of islanded microgrids. Background Technology

[0002] An islanded microgrid specifically refers to a microgrid system that cannot be connected to the main grid and needs to rely on local power sources to maintain its power supply independently. Typical application scenarios include island microgrids, remote station microgrids, offshore platform microgrids, and independent field station microgrids—scenarios where power cannot be supplied by the main grid. An islanded microgrid typically includes renewable energy generation units, diesel generator sets, energy storage systems, and various loads. Renewable energy generation units can be wind power, solar power, or other fluctuating power sources; diesel generator sets offer strong controllability and are one of the most important dispatchable power sources in an islanded microgrid; energy storage systems can be used to smooth out fluctuations, reduce instantaneous gaps, or absorb surplus energy in short periods; loads are typically divided into critical loads and general loads according to their importance.

[0003] Isolated microgrids cannot adjust the gap between local generation capacity and load demand by purchasing electricity from the main grid or feeding back electricity. Their long-term stable and continuous operation capability mainly depends on the dispatch resources for sustainable local supply. For isolated microgrids equipped with diesel generators, although diesel generators have strong output controllability, their continuous generation capacity is directly constrained by fuel inventory.

[0004] In islanded microgrid dispatching schemes, diesel engines are typically treated as ordinary dispatchable power sources, with fuel consumption reflected as an economic cost. Refueling often relies on manual experience, maintenance cycles, or planned transportation arrangements, primarily focusing on short-term dispatch optimization around load forecasting, renewable energy forecasting, power balance constraints, and energy storage state of charge constraints. However, these schemes lack systematic modeling of dynamic changes in fuel inventory, fuel safety margins, and the timing and amount of refueling. They struggle to assess fuel shortage risks based on future load and renewable energy output changes and lack an integrated linkage mechanism with diesel engine dispatching, energy storage regulation, and load management. Summary of the Invention

[0005] The purpose of this application is to provide an optimized scheduling and fuel replenishment decision-making method for isolated microgrids, which can realize the joint optimization of energy storage regulation, diesel engine scheduling, load management and fuel replenishment actions of isolated microgrids, and achieve synergistic optimization of energy storage and diesel engines.

[0006] Firstly, this application provides a method for optimized scheduling and refueling decision-making in islanded microgrids, the method comprising: For any given scheduling time, Based on the status information of the islanded microgrid at the current scheduling time and the historical data within the historical time window before the current scheduling time, the load forecast sequence and renewable energy output forecast sequence of the islanded microgrid in the future time window including the current scheduling time are determined. The net load gap assessment value within the future time window is determined based on the load forecast sequence and the renewable energy output forecast sequence. Based on the net load gap assessment value within the future time window and the measured energy storage state of charge at the current dispatch time, the energy storage support power assessment value and the minimum diesel demand power forecast value within the future time window are determined. Based on the energy storage support power assessment value and the predicted fuel inventory in the future time window, the predicted equivalent energy inventory in the future time window is determined. Based on the minimum diesel demand power forecast value in the future time window, the dynamic safety stock boundary in the future time window is determined. Based on the dynamic safety stock boundary and the predicted equivalent energy inventory in the future time window, the reference state of charge in the future time window is determined. An objective function is constructed to characterize the comprehensive operating cost of an islanded microgrid within a future time window. A mixed-integer linear programming joint optimization model is then built using the objective function, constraints, and scheduling parameters. The constraints include power balance constraints, diesel engine operation constraints, fuel inventory constraints, energy storage constraints, refueling delay constraints, refueling quantity constraints, and load reduction constraints. The fuel inventory constraints are determined based on the dynamic safety stock boundary and equivalent energy inventory within the future time window, and the energy storage constraints are determined based on the reference state of charge within the future time window. The scheduling parameters include diesel engine control parameters, energy storage control parameters, load control parameters, and refueling control parameters. Based on the state information at the current scheduling moment, the mixed integer linear programming joint optimization model is solved to obtain the optimal scheduling parameters at the current scheduling moment, and the islanded microgrid is controlled using the optimal scheduling parameters.

[0007] In some embodiments, the status information includes the current measured remaining fuel quantity, the current diesel engine start / stop status, the current actual diesel engine output, the current energy storage system state of charge, the current actual renewable energy output, the current measured total load, and the current status of the on-the-go refueling task. The future time window includes H time periods t. Determining the energy storage support power assessment value and the minimum diesel demand power prediction value within the future time window includes: For any time period t, The net load gap assessment value for time period t is determined by the difference between the load forecast and the renewable energy output forecast for that time period t. Based on the measured state of charge of energy storage, rated capacity of energy storage, lower limit of basic allowable limit of energy storage and discharge efficiency at the current scheduling time, determine the total amount of energy that can be released by energy storage at the current scheduling time. Based on the total energy storage capacity that can be released at the current scheduling time, the maximum allowable discharge power of energy storage, and the net load gap assessment value for time period t, the energy storage support power assessment value for time period t is determined. The minimum diesel demand power forecast for time period t is determined by the difference between the energy storage support power assessment value and the net load gap. The estimated fuel consumption during the period t is determined by using the predicted minimum diesel power demand for the time period t and the start-up indication during the period t when operating at the minimum diesel power demand.

[0008] In some embodiments, the predicted equivalent energy inventory for the future time window is determined based on the energy storage support power assessment value and the predicted fuel inventory for the future time window; the dynamic safety stock boundary for the future time window is determined based on the minimum diesel demand power forecast value for the future time window; and the reference state of charge for the future time window is determined based on the dynamic safety stock boundary and the predicted equivalent energy inventory for the future time window, including: For any time period t, Based on the predicted fuel inventory of the previous time period t-1, the estimated fuel consumption during time period t at the minimum diesel demand power, and the known fuel replenishment amount that has been triggered before the current scheduling time k and is expected to arrive in time period t, the predicted fuel inventory for time period t is determined using the fuel inventory recursion formula. Based on the energy storage support power assessment value for time period t, the total amount of energy that can be released from energy storage at the current scheduling time, and the equivalent coefficient that a unit of fuel can be converted into electricity, determine the energy storage equivalent fuel inventory assessment value for time period t. The estimated equivalent fuel oil inventory for the time period t and the predicted fuel oil inventory for the time period t are used to determine the predicted equivalent energy inventory for the time period t. Based on the assessed fuel consumption during the period t at the minimum diesel demand power and the safety margin term for that period t, the dynamic safety stock boundary for that period t is determined. The safety margin term for that period t is determined based on the load forecast error conversion, renewable energy forecast error conversion, critical load power, and refueling delay for that period t. Based on the dynamic safety stock boundary and the predicted equivalent energy stock for time period t, the inventory risk index for time period t is determined. Based on the inventory risk index, low-risk reference state of charge, high-risk reference state of charge, and saturation function of the risk mapping interval parameter for time period t, the reference state of charge for time period t is determined.

[0009] In some embodiments, Use the following fuel inventory recursive formula to determine the predicted fuel inventory for the t-th time period. for: ; ; in, This represents the initial conditions of the formula. This indicates the actual remaining fuel quantity at the current dispatch time. This represents the predicted fuel inventory for the previous time period t-1; This represents the estimated fuel consumption during the t-th time period when operating at the minimum diesel power requirement. This represents the known amount of fuel replenishment that has been triggered before the current scheduling time k and is expected to arrive in time period t. The remaining releasable energy from the stored energy in the t-th time period is determined using the following formula. ; ; in, This indicates the total amount of electricity that can be released from energy storage at the time of dispatch. This represents the energy storage support power assessment value within the t-th time period; The energy storage equivalent fuel inventory assessment value for the t-th time period is determined using the following formula. ; ; in, This represents the equivalent energy that a unit of diesel fuel can be converted into electrical energy. The predicted equivalent energy inventory for the t-th time period is determined using the following formula. ; ; The dynamic safety stock boundary for the t-th time period Represented as; ; ; in, This indicates the lower limit of the physical safety of the fuel depot. This represents the safety margin term for the t-th time period. This represents the number of discrete time intervals corresponding to the refueling delay. This indicates the actual time length corresponding to the refueling delay. This represents the estimated fuel consumption obtained when operating at the minimum diesel power requirement during the i-th time period. , , This represents the conversion factor. This represents the adjusted load forecast error for the t-th time period. This represents the calculated error in renewable energy forecasts for the t-th time period. This represents the critical load power in the t-th time period; The following formula is used to determine the inventory risk indicator for the t-th time period. ; ; in, Represents small positive numbers; The reference state of charge in the t-th time period Represented as; ; in, Indicates a low-risk reference state of charge. Indicates a high-risk reference state of charge. and This represents the risk mapping interval parameter. This represents a saturation function.

[0010] In some embodiments, an objective function is constructed to characterize the comprehensive operating cost of the islanded microgrid within a future time window. A mixed-integer linear programming joint optimization model is then constructed using the objective function, constraints, and scheduling parameters, including: The scheduling parameters include diesel engine control parameters, energy storage control parameters, load control parameters, and fuel replenishment control parameters. The diesel engine control parameters include diesel engine start-up variables, diesel engine start-up state variables, and diesel engine scheduling output. The energy storage control parameters include energy storage charging power, energy storage discharging power, and energy storage state of charge at the end of the t-th time period. The load control parameters include general load reduction power. The fuel replenishment control parameters include fuel replenishment trigger variables, fuel replenishment arrival state variables, and the arrival amount of the new fuel replenishment task triggered by the joint optimization model at the t-th time period. The comprehensive operating costs of an isolated microgrid within a future time window include fuel consumption costs, diesel engine start-up costs, load shedding penalties, energy storage state-of-charge deviation costs, and energy storage usage costs. To minimize these comprehensive operating costs, an objective function is constructed. ; ; ; ; in, This represents the unit fuel cost coefficient. This represents the diesel engine fuel consumption in the t-th time period. This represents the dispatch output of the diesel engine in the t-th time period. This represents the diesel engine start-up status variable during the t-th time period. This represents the cost coefficient for a single start-up of a diesel engine. This represents the diesel engine starting variable during the t-th time period. This represents the general load reduction penalty factor. This represents the typical load reduction power during the t-th time period. This indicates the interval between two adjacent scheduling times. This represents the weighting coefficient for the deviation of the state of charge. This represents the energy storage state of charge deviation in the t-th time period. This represents the energy storage usage cost coefficient. This represents the energy storage charging power in the t-th time period. This represents the energy storage discharge power during the t-th time period.

[0011] In some embodiments, the method further includes: Remaining fuel quantity after the t-th time period ends for; ; ; ; in, This represents the amount of fuel remaining after the (t-1)th time period. This represents the initial conditions of the formula. This represents the total amount of fuel replenished in the t-th time period. This represents the arrival amount of the newly triggered refueling task output by the joint optimization model in the t-th time period; Using the following recursive formula for the state of charge of energy storage, determine the state of charge of energy storage at the end of the t-th time period. : ; ; middle, Indicates the rated capacity of energy storage. Indicates charging efficiency. Indicates discharge efficiency. This represents the initial conditions of the formula. This indicates the measured state of charge of the energy storage at the current scheduling moment. This represents the state of charge of the energy storage at the end of the (t-1)th time period.

[0012] In some embodiments, the constraints are specifically: The power balance constraint satisfies the following formula: ; in, This represents the dispatched output of the diesel engine in the t-th time period. This represents the energy storage discharge power in the t-th time period. This represents the typical load reduction power during the t-th time period. This represents the predicted renewable energy output for the t-th time period. This represents the load forecast for the t-th time period. This represents the energy storage charging power during the t-th time period; The operating constraints of a diesel engine satisfy the following formula: ; ; in, This represents the dispatch output of the diesel engine in the t-th time period. This indicates the minimum permissible output of the diesel engine. This indicates the maximum permissible output of the diesel engine. This represents the diesel engine's start-up state variable during the t-th time period. This represents the diesel engine start-up state variable during the (t-1)th time period. This represents the diesel engine starting variable during the t-th time period; Fuel inventory constraints satisfy the following formula: ; ; in, This represents the amount of fuel remaining after the t-th time period ends. This represents the predicted equivalent energy inventory for the t-th time period. Indicates minimum fuel inventory. Indicates the maximum capacity of the oil storage unit; Energy storage constraints satisfy the following formula: ; ; in, Indicates the maximum allowable charging power of the energy storage. Indicates the maximum allowable discharge power of the energy storage. This represents the energy storage charging power in the t-th time period. This represents the energy storage discharge power during the t-th time period; The refueling delay constraint is: ; in, This represents the state variable indicating the arrival of refueling in the (t+Loil)th time period. This represents the refueling trigger variable in the t-th time period; The fuel replenishment constraint satisfies the following formula: ; in, This represents the arrival amount of the newly triggered refueling task output by the joint optimization model in the t-th time period. This indicates the lower limit of a single refueling operation. This indicates the maximum amount of fuel that can be added in a single transaction. This represents the state variable indicating the arrival of refueling in the t-th time period; The load reduction constraint satisfies the following formula: ; in, This represents the typical load reduction power during the t-th time period. This represents the maximum power limit of the general load that can be reduced during the t-th time period.

[0013] In some embodiments, based on the state information at the current scheduling moment, the mixed-integer linear programming joint optimization model is solved to obtain the optimal scheduling parameters at the current scheduling moment, and the islanded microgrid is controlled using the optimal scheduling parameters, including: Based on the state information at the current scheduling moment, the mixed integer linear programming joint optimization model is solved to obtain the optimal scheduling parameter sequence within the future time window, wherein the future time window includes H time periods, and the optimal scheduling parameter sequence includes the optimal scheduling parameters for the H time periods. Obtain the optimal scheduling parameters corresponding to the current scheduling time from the optimal scheduling parameter sequence, and use the optimal scheduling parameters corresponding to the current scheduling time to perform control on the islanded microgrid.

[0014] In some embodiments, obtaining the optimal scheduling parameters corresponding to the current scheduling time from the optimal scheduling parameter sequence, and performing control on the islanded microgrid using the optimal scheduling parameters corresponding to the current scheduling time, includes: Based on the optimal diesel engine start-up variables, optimal diesel engine start-up status variables, and optimal diesel engine dispatch output corresponding to the current dispatch time, start-up or stop status, start-up action, and dispatch output command are issued to the diesel engine. Based on the optimal energy storage charging power variable and the optimal energy storage charging power variable corresponding to the current scheduling time, charging power and discharging power commands are issued to the energy storage system. When the general load reduction power is greater than 0, general load reduction control is performed on the load side. If the refueling trigger variable corresponding to the current scheduling time is 1, a refueling command is triggered at the current scheduling time, and the newly triggered refueling task is added to the in-transit refueling task status. Based on the islanded microgrid after execution control, the status information of the next scheduling moment is collected.

[0015] Secondly, this application provides an islanded microgrid optimized scheduling and refueling decision-making system, the system comprising: The data acquisition device is used to obtain the status information of the islanded microgrid at any given scheduling time. The prediction module is used to determine the load prediction sequence and renewable energy output prediction sequence of the islanded microgrid in the future time window, including the current scheduling time, based on the acquired state information of the islanded microgrid at the current scheduling time and historical data within the historical time window before the current scheduling time. The load gap and demand assessment module is used to determine the net load gap assessment value within the future time window based on the load forecast sequence and the renewable energy output forecast sequence, and to determine the energy storage support power assessment value and the minimum diesel demand power forecast value within the future time window based on the net load gap assessment value within the future time window and the measured energy storage state of charge at the current dispatch time. The calculation module is used to determine the predicted equivalent energy inventory in the future time window based on the energy storage support power assessment value and the predicted fuel inventory in the future time window, to determine the dynamic safety stock boundary in the future time window based on the minimum diesel demand power forecast value in the future time window, and to determine the reference state of charge in the future time window based on the dynamic safety stock boundary and the predicted equivalent energy inventory in the future time window. The optimization model module is used to construct an objective function to characterize the comprehensive operating cost of the islanded microgrid within a future time window. A mixed-integer linear programming joint optimization model is constructed based on the objective function, constraints, and scheduling parameters. The constraints include power balance constraints, diesel engine operation constraints, fuel inventory constraints, energy storage constraints, refueling delay constraints, refueling quantity constraints, and load reduction constraints. The fuel inventory constraints are determined based on the dynamic safety stock boundary and equivalent energy inventory within the future time window, and the energy storage constraints are determined based on the reference state of charge within the future time window. The scheduling parameters include diesel engine control parameters, energy storage control parameters, load control parameters, and refueling control parameters. The control module is used to solve the mixed integer linear programming joint optimization model based on the state information at the current scheduling time, obtain the optimal scheduling parameters at the current scheduling time, and perform control on the islanded microgrid with the optimal scheduling parameters.

[0016] This application, based on load forecasting and renewable energy output forecasting within a future time window, determines the net load gap, the energy storage's alternative energy supply capacity, and the minimum diesel demand to construct a dynamic safety stock boundary. It then establishes a mixed-integer linear programming joint optimization model that considers diesel engine start-stop, diesel engine output, energy storage charging and discharging, general load reduction, and refueling actions, directly outputting optimal scheduling parameters for corresponding control of the microgrid system. Fuel inventory is defined as a state variable in the model, and the available energy of the energy storage system is converted into an equivalent fuel inventory, which, together with the actual fuel inventory, constitutes the system's equivalent energy inventory. The available energy of the energy storage system is included in the long-term energy supply capacity assessment. Based on this equivalent energy inventory, future energy supply capacity assessment, inventory boundary construction, and scheduling decisions are performed, thereby achieving coordinated optimization of energy storage and diesel engines. A closed-loop control is established between load forecasting, inventory assessment, diesel engine scheduling, energy storage adjustment, and refueling determination. An inventory risk index is constructed using the dynamic safety stock boundary, and then the inventory risk generates a reference state of charge and a lower bound for strategic reserves for energy storage, upgrading energy storage behavior from passively satisfying constraints to actively adjusting around inventory risk. By using refueling actions as model decision variables, a closed-loop operation mechanism of "prediction - inventory assessment - energy storage guidance - dispatch - refueling - rolling update" is established. Attached Figure Description

[0017] Figure 1 This is a first flowchart of the islanded microgrid optimization scheduling and refueling decision-making method provided in the embodiments of this application; Figure 2 A second flowchart is provided for the islanded microgrid optimized scheduling and refueling decision method in the embodiments of this application; Figure 3 The third flowchart is provided for the islanded microgrid optimization scheduling and refueling decision-making method in the embodiments of this application; Figure 4 The fourth flowchart is provided for the islanded microgrid optimization scheduling and refueling decision-making method in the embodiments of this application; Figure 5 The system block diagram of the islanded microgrid optimization scheduling and refueling decision system provided in the embodiments of this application is shown. Detailed Implementation

[0018] The present application will be described in detail below with reference to the specific embodiments shown in the accompanying drawings. However, these embodiments do not limit the present application. Any structural, methodological, or functional modifications made by those skilled in the art based on these embodiments are included within the protection scope of the present application.

[0019] Please refer to Figure 1 This application provides a method for optimized scheduling and refueling decision-making in an islanded microgrid, the method comprising: S101, For any scheduling time, based on the acquired state information of the islanded microgrid at the current scheduling time and the historical data within the historical time window before the current scheduling time, determine the load forecast sequence and renewable energy output forecast sequence of the islanded microgrid within the future time window including the current scheduling time. S102, Based on the load forecast sequence and the renewable energy output forecast sequence, determine the net load gap assessment value within the future time window, and based on the net load gap assessment value within the future time window and the measured energy storage state of charge at the current dispatch time, determine the energy storage support power assessment value and the minimum diesel demand power forecast value within the future time window. S103, Based on the energy storage support power assessment value and the predicted fuel inventory in the future time window, determine the predicted equivalent energy inventory in the future time window; Based on the minimum diesel demand power forecast value in the future time window, determine the dynamic safety stock boundary in the future time window; Based on the dynamic safety stock boundary and the predicted equivalent energy inventory in the future time window, determine the reference state of charge in the future time window. S104, construct an objective function to characterize the comprehensive operating cost of the islanded microgrid in the future time window, and construct a mixed integer linear programming joint optimization model with the objective function, constraints and scheduling parameters; S105, based on the state information at the current scheduling moment, solves the mixed integer linear programming joint optimization model to obtain the optimal scheduling parameters at the current scheduling moment, and performs control on the islanded microgrid with the optimal scheduling parameters.

[0020] At the current scheduling moment, the status information of the isolated microgrid is collected and acquired. This status information includes the current measured remaining fuel quantity, current diesel engine start / stop status, current actual diesel engine output, current energy storage system state of charge, current actual renewable energy output, current measured total load, and the status of currently en route refueling tasks. For example, the interval between two adjacent scheduling moments is 15 minutes. The energy storage state of charge characterizes the energy storage inventory status. The status of currently en route refueling tasks represents refueling tasks that have been triggered before the current scheduling moment but have not yet arrived, including the expected arrival time, expected arrival quantity, and current execution status. Refueling actions are incorporated into the joint optimization model. At each scheduling moment, in addition to re-collecting the fuel inventory, diesel engine status, energy storage status, and load status, the status of currently en route refueling tasks must also be updated synchronously to ensure the continuity and executability of cross-window refueling actions in rolling control.

[0021] For example, the current measured remaining fuel quantity can be obtained by collecting the liquid level signal from the liquid level sensor and then calculated by the microgrid monitoring system or on-site acquisition device in conjunction with the fuel tank calibration relationship. The current diesel engine start / stop status, current actual diesel engine output, current energy storage system state of charge, current actual renewable energy output, and current measured total load can be obtained in real time by the microgrid monitoring system or on-site acquisition device; the status of in-transit refueling tasks can be obtained in real time by the operation and maintenance scheduling system, transportation scheduling system, or refueling task registration module.

[0022] The constraints include power balance constraints, diesel engine operation constraints, fuel inventory constraints, energy storage constraints, refueling delay constraints, refueling quantity constraints, and load reduction constraints. Fuel inventory constraints are determined based on the dynamic safety stock boundary and equivalent energy inventory within the future time window, while energy storage constraints are determined based on the reference state of charge within the future time window. Scheduling parameters include diesel engine control parameters, energy storage control parameters, load control parameters, and refueling control parameters. Diesel engine control parameters include diesel engine start-up variables, diesel engine start-up state variables, and diesel engine dispatch output. Energy storage control parameters include energy storage charging power, energy storage discharging power, and the energy storage state of charge at the end of time period t. Load control parameters include general load reduction power. Refueling control parameters include refueling trigger variables, refueling arrival state variables, and the arrival quantity of the newly triggered refueling task output by the joint optimization model in time period t.

[0023] In some embodiments, time series methods, machine learning methods, neural network methods, or combinations thereof are used to obtain load forecast sequences and renewable energy output forecast sequences within future time windows. The renewable energy output forecast sequence refers to a sequence composed of output data within the current scheduling time, and the load forecast sequence refers to a sequence composed of load data within the current scheduling time. Taking a Long Short-Term Memory (LSTM) network model as an example, this LTM network model is a pre-trained model. Historical data includes historical operating data, external environment data, time characteristic data, and microgrid internal operating plans. Historical operating data may include load curves and renewable energy output curves within historical time windows. External environment data may include solar irradiance, wind speed, temperature, weather type, etc. Time characteristic data may include time periods, dates, weekdays / holidays, electricity consumption plans, etc. Microgrid internal operating plans may include known production loads, residential loads, and equipment start-up and shutdown plans. The current scheduling time's state information and historical data are used to construct the model's input feature matrix. The LTM network model outputs load forecast sequences and renewable energy output forecast sequences within future time windows, providing a basis for subsequent steps such as obtaining net load gaps, minimum diesel demand assessments, and safety stock boundaries.

[0024] In this embodiment, based on load forecasting and renewable energy output forecasting within a future time window, the net load gap, energy storage's alternative energy supply capacity, and minimum diesel demand are determined to construct a dynamic safety stock boundary. A mixed-integer linear programming joint optimization model is established, considering diesel engine start-stop, diesel engine output, energy storage charging and discharging, general load reduction, and refueling actions. This directly outputs the diesel engine scheduling scheme, energy storage regulation scheme, load management scheme, refueling trigger scheme, refueling arrival scheme, and future stock trajectory within the future time window. Corresponding control is then implemented on the microgrid system. Fuel inventory is defined as the model's state variable, and the available energy of the energy storage system is converted into equivalent fuel inventory. This, together with the actual fuel inventory, constitutes the system's equivalent energy inventory, forming a multi-energy inventory system. This allows energy storage to participate not only in short-term power balance but also in long-term system energy supply capacity assessment and scheduling decisions. Based on this equivalent energy inventory, future energy supply capacity assessment, inventory boundary construction, and scheduling decisions are performed, thereby achieving synergistic optimization of energy storage and diesel engines. A closed-loop control is established between load forecasting, inventory assessment, diesel engine scheduling, energy storage regulation, and refueling determination. By constructing an inventory risk indicator using a dynamic safety stock boundary, and then generating a reference state of charge and a lower bound for strategic reserves based on the inventory risk, energy storage behavior is upgraded from passively satisfying constraints to actively adjusting around inventory risk. Refueling actions are used as a model decision variable to establish a closed-loop operation mechanism of "prediction—inventory assessment—energy storage guidance—scheduling—refueling—rolling update".

[0025] In some embodiments, such as Figure 2 As shown, the future time window includes H time periods t. Based on the net load gap and energy storage state of charge within the future time window, the available energy storage capacity and minimum diesel demand power forecast within the future time window are determined, including: S201, For any time period t, determine the net load gap assessment value for that time period t by using the difference between the load forecast and the renewable energy output forecast for that time period t; S202. Based on the measured state of charge of energy storage, rated capacity of energy storage, lower limit of basic allowable capacity of energy storage and discharge efficiency at the current scheduling time, determine the total amount of energy storage that can be released at the current scheduling time. S203. Based on the total energy storage that can be released at the current scheduling time, the maximum allowable discharge power of energy storage, and the net load gap assessment value for the time period t, determine the energy storage support power assessment value for the time period t. S204. The minimum diesel demand power forecast for time period t is determined by the difference between the energy storage support power assessment value and the net load gap for time period t. S205, using the predicted minimum diesel power demand for time period t and the start-up indication for time period t when operating at minimum diesel power demand, determine the estimated fuel consumption for time period t when operating at minimum diesel power demand.

[0026] For example, load forecasting sequence Represented as: ; Renewable Energy Output Forecast Series Represented as: ; Where H represents the duration of the future time window, that is, the future time window has H time periods t, and k represents the current scheduling time.

[0027] For example, the net load gap assessment value for the t-th time period The calculation formula is: ; in, This represents the load forecast for the t-th time period. This represents the predicted renewable energy output for the t-th time period; The above calculation formula can be further expressed as: ; The above formula represents the output shortfall that an islanded microgrid needs to fill with dispatchable power sources during time period t, without considering the specific dispatch actions of diesel engines and energy storage. When this occurs, it means that the predicted renewable energy output for that time period t is sufficient to cover the load demand, and in principle, there is no power shortage in the system that must be additionally handled by diesel engines; when The time interval t indicates that there is a power shortage that needs to be shared by both diesel engines and energy storage.

[0028] For example, to uniformly assess the contribution of energy storage to long-term energy supply capacity, the total amount of electricity that energy storage can release at the time of dispatch is... The calculation formula is: ; in, Indicates the rated capacity of energy storage. This indicates the measured state of charge of the energy storage at the current scheduling moment. This indicates the lower limit of the allowable limit for energy storage infrastructure. This represents the discharge efficiency. The basic allowable lower limit characterizes the physical safety boundary of energy storage, while the inventory risk-driven reference state of charge and reserve constraints calculated in subsequent steps characterize the strategic power supply requirements at the operational level. The basic allowable lower limit can be set according to the physical characteristics of the energy storage battery at the time of manufacture, such as a protection threshold to prevent damage from deep discharge, typically 10% or 20%.

[0029] For example, the energy storage-supported power assessment value for the t-th time period. The calculation formula is: ; ; ; in, Indicates the maximum allowable discharge power of the energy storage. This indicates the interval between two adjacent scheduling times.

[0030] For example, the minimum diesel power demand forecast for the t-th time period. Represented as: ; For example, the estimated fuel consumption during the t-th time period when operating at the minimum diesel power requirement. ; ; in, This represents the fuel consumption coefficient related to diesel engine output. This represents the basic fuel consumption coefficient for diesel engines at startup. This represents the predicted minimum diesel power demand for the t-th time period. This represents the start-up state variable during the t-th time period when the generator is operating at the minimum diesel power demand. hour, Select 1 if the value is 1, otherwise select 0. This indicates the interval between two adjacent scheduling times.

[0031] The minimum diesel demand power forecast is an equivalent demand used for subsequent inventory assessment, for calculating the minimum fuel demand and constructing the dynamic safety stock boundary, and does not constitute a rigid constraint on the actual dispatch sequence. This application incorporates the substitution effect of energy storage on diesel generator output into the subsequent joint optimization model, which determines the actual charging and discharging power and diesel engine output.

[0032] In some embodiments, such as Figure 3 As shown, the predicted equivalent energy inventory for the future time window is determined based on the energy storage support power assessment value and the predicted fuel inventory for the future time window. The dynamic safety stock boundary for the future time window is determined based on the minimum diesel demand power forecast value for the future time window. The reference state of charge for the future time window is determined based on the dynamic safety stock boundary and the predicted equivalent energy inventory, including: S301, for any time period t, based on the predicted fuel inventory of the previous time period t-1, the estimated fuel consumption during time period t at the minimum diesel demand power, and the known fuel replenishment amount that has been triggered before the current scheduling time and is expected to arrive in time period t, the predicted fuel inventory for time period t is determined using the fuel inventory recursion formula. S302, based on the energy storage support power assessment value for time period t, the total energy that can be released from the energy storage at the current scheduling time, and the equivalent coefficient that a unit of fuel can be converted into electrical energy, determine the energy storage equivalent fuel inventory assessment value for time period t. S303, the predicted equivalent energy inventory for time period t is determined by the energy storage equivalent fuel inventory assessment value and the predicted fuel inventory for time period t. S304, Based on the assessed fuel consumption during operation at the minimum diesel demand power for time period t and the safety margin term for time period t, determine the dynamic safety stock boundary for time period t, wherein the safety margin term for time period t is determined based on the load forecast error conversion amount, renewable energy forecast error conversion amount, critical load power and refueling delay for time period t. S305, Based on the dynamic safety stock boundary and the predicted equivalent energy stock for time period t, determine the inventory risk index for time period t. S306. Based on the inventory risk index, low-risk reference state of charge, high-risk reference state of charge, and saturation function of the risk mapping interval parameter for the time period t, determine the reference state of charge for the time period t.

[0033] For example, the predicted fuel inventory for the t-th time period is determined using the following fuel inventory recursive formula. : ; in, This represents the initial conditions of the formula. This indicates the actual remaining fuel quantity at the current dispatch time. This represents the predicted fuel inventory for the previous time period t-1; This represents the estimated fuel consumption during the t-th time period when operating at the minimum diesel power requirement. This represents the known amount of fuel replenishment that has been triggered before the current scheduling time k and is expected to arrive within time period t. It is based on the measured remaining fuel quantity at the current scheduling time k. as initial value Using the above recursive formula, the predicted fuel inventory for each time period t can be obtained sequentially. If there are no known refueling missions en route, then It is 0.

[0034] For example, the remaining releasable energy storage capacity in the t-th time period is determined using the following formula. ; ; in, This indicates the total amount of electricity that can be released from energy storage at the time of dispatch. This represents the energy storage support power assessment value during the t-th time period.

[0035] The energy storage equivalent fuel inventory assessment value for the t-th time period is determined using the following formula. : ; in, It represents the equivalent coefficient of electrical energy that can be converted from a unit of diesel fuel, and represents the amount of electricity that can be generated per unit of fuel.

[0036] For example, the predicted equivalent energy inventory for the t-th time period is determined using the following formula. : ; The above formula indicates that the comprehensive energy supply resources in the t-th time period include not only the remaining fuel oil in that time period t, but also the equivalent fuel oil inventory that can replace diesel power supply during that time period t.

[0037] The dynamic safety stock boundary represents the minimum inventory level that should be maintained to cover subsequent energy supply risks in time period t. For example, the dynamic safety stock boundary in time period t... Represented as: ; ; in, This indicates the lower limit of the physical safety of the fuel depot. This represents the safety margin term for time period t. This represents the number of discrete time intervals corresponding to the refueling delay, that is, the number of time intervals required from triggering the refueling command to the arrival of the refueled vehicle. This indicates the actual time length corresponding to the refueling delay. This represents the estimated fuel consumption at the current scheduling time k, during the i-th time period when operating at the minimum diesel demand power. The safety margin term is used to cover uncertainties such as load forecasting errors, renewable energy forecasting errors, critical load supply requirements, fuel consumption model errors, and refueling delays. , , This represents the conversion factor. This represents the adjusted load forecasting error for time period t. This represents the adjusted amount of the renewable energy forecast error. Indicates the critical load power. For example, , , The values ​​can be 1.0, 1.0 and 1.2 respectively.

[0038] In one embodiment, and The forecast error can be set to a variable amount over future time periods. For example, the forecast error can be adjusted for different future time periods based on real-time meteorological data, historical load data, load forecasting model outputs, and renewable energy output forecasting model outputs. In one embodiment, to reduce the difficulty of data acquisition and model computation complexity, the forecast error can be set to a uniform fixed value for each time period within the future time window. It can be set as a preset percentage of the rated load or the corresponding predicted load value, such as taking 5% of the maximum predicted load value; It can be set as a preset proportion of the installed capacity of renewable energy or the corresponding predicted value of renewable energy output. It can be determined by the sum of the load capacities that the microgrid must guarantee, and is a constant value, such as the total power of radar and communication base stations in an island microgrid of 50kW.

[0039] To enable energy storage to participate more in replacing diesel power supply when inventory is ample and to proactively retain emergency power supply capacity when inventory is tight, a reference state of charge (SOC) is generated for each time period based on inventory risk, and this SOC is used as a known guiding parameter for the joint optimization model. The reference SOC is not the actual state of the energy storage system, but rather a target energy level set according to the current inventory risk level, used to guide the charging and discharging behavior of the energy storage system during the optimization process.

[0040] For example, the inventory risk indicator for the t-th time period is determined using the following formula. ; ; in, Represents small positive numbers. This represents the predicted equivalent energy inventory within the t-th time period.

[0041] To avoid abrupt changes in the reference state of charge (SBC) between adjacent rolling moments, a first-order smoothing mechanism is introduced, ensuring continuous temporal variation of the SBC, thereby suppressing diesel engine start-stop oscillations and energy storage charging / discharging oscillations. Piecewise linear or saturated mapping methods can be used to transform the inventory risk indicator into the SBC. The SBC is defined as follows: Reference State of Charge in the t-th time period. for: ; in, This indicates a low-risk reference state of charge, which can be set to, for example, 20%. This indicates a high-risk reference state of charge, which can be set to, for example, 90%. and This represents the risk mapping interval parameter, which can be set to 1.0 and 1.5 respectively. The saturation function is represented as: .

[0042] In this embodiment, the inventory risk index is determined by the dynamic safety stock boundary within the future time window, and the reference state of charge within the future time window is further determined, forming a parameter generation chain. The objective function and constraints of the subsequent joint optimization model are all related to the reference state of charge, thus forming a closed-loop mechanism of "inventory risk - reference state of charge - energy storage behavior - inventory evolution".

[0043] The set scheduling parameters are used as decision variables in the mixed-integer linear programming joint optimization model. The scheduling parameters include diesel engine control parameters, energy storage control parameters, load control parameters, and fuel replenishment control parameters.

[0044] Diesel engine control parameters include diesel engine starting variables. Diesel engine start-up state variables And the dispatching output of diesel engines Energy storage control parameters include energy storage charging power. Energy storage and discharge power and the state of charge of energy storage at the end of the t-th time period Load control parameters include general load reduction power. The fuel replenishment control parameters include the fuel replenishment trigger variable. Refueling reaches the state variable The amount of fuel replenishment corresponding to the new trigger. .

[0045] In some embodiments, an objective function is constructed to characterize the comprehensive operating cost of the islanded microgrid within a future time window. A mixed-integer linear programming joint optimization model is then constructed using the objective function, constraints, and scheduling parameters, including: The comprehensive operating costs of an isolated microgrid within a future time window include fuel consumption costs, diesel engine start-up costs, load shedding penalties, energy storage state-of-charge deviation costs, and energy storage usage costs. To minimize these comprehensive operating costs, an objective function is constructed. ; ; ; ; To maintain the overall linearity of the model, the following auxiliary variable constraints can be used: ; ; ; in, This represents the unit fuel cost coefficient. This represents the diesel engine fuel consumption in the t-th time period. This represents the dispatch output of the diesel engine in the t-th time period. This represents the diesel engine start-up status variable during the t-th time period. This represents the cost coefficient for a single start-up of a diesel engine. This represents the diesel engine starting variable during the t-th time period. This represents the general load reduction penalty factor. This represents the typical load reduction power during the t-th time period. This indicates the interval between two adjacent scheduling times. This represents the weighting coefficient for the deviation of the state of charge. This represents the energy storage state of charge deviation in the t-th time period. This represents the energy storage usage cost coefficient. This represents the energy storage charging power in the t-th time period. This represents the energy storage discharge power during the t-th time period. The term used to linearly characterize the deviation of the state of charge from the reference state of charge within the t-th time interval can be understood as applying a nonlinear "absolute value" penalty term (i.e., ... The auxiliary decision variable is introduced when the problem is transformed into a linear programming problem. Its final value is automatically calculated by the solver of the mixed integer linear programming joint optimization model in the subsequent steps, under the premise of satisfying the constraints, in order to minimize the objective function.

[0046] In the objective function described above, the fuel consumption cost term is used to reduce the fuel consumption level of diesel engines; the diesel engine start-up cost term is used to avoid frequent start-stop of diesel engines, reduce mechanical wear and cold start frequency; the load reduction penalty term is used to mathematically embed supply requirements into the model; the energy storage state of charge deviation cost term is used to enable energy storage to maintain higher energy reserves when inventory risk is high, and to increase the initiative of energy storage to participate in power supply when inventory risk is low; and the energy storage usage cost term suppresses ineffective or excessively frequent charging and discharging behavior.

[0047] Use the following fuel inventory recursive formula to determine the amount of fuel remaining at the end of the t-th time period. for: ; ; ; ; in, Let represent the remaining fuel quantity after the (t-1)th time period in a rolling optimization cycle initiated at the current scheduling time k. This represents the initial condition for the recursive formula, defined as the measured remaining fuel quantity at the current scheduling time k. , This represents the total amount of refueling that arrives and can be counted in inventory during the t-th time period, including the known amount of refueling that is en route and the amount of refueling that is newly triggered. This indicates the amount of fuel that has been replenished upon triggering the new fuel replenishment. This represents the diesel engine fuel consumption in the t-th time period. This represents the dispatch output of the diesel engine in the t-th time period. This represents the diesel engine start-up status variable during the t-th time period.

[0048] In the above recursive formula, As a future state variable in the optimization model, it represents the state trajectory of the remaining fuel quantity within a future time window. As the initial condition for the recursive formula, within the time interval t=1, based on Optimize model output as well as The above formula is used to calculate the result. By analogy, the remaining fuel amount in each time period from t=2 to t=H can be obtained.

[0049] For example, the energy storage state of charge at the end of the t-th time period is determined using the following recursive formula. : ; ; in, Indicates the rated capacity of energy storage. Indicates charging efficiency. Indicates discharge efficiency. This represents the initial condition for the energy storage state of charge, defined as the measured state of charge of the energy storage at the current scheduling time K. , This represents the state of charge of the energy storage at the end of the (t-1)th time period.

[0050] In the above recursive formula, As a future state variable in the optimization model, it is used to represent the state trajectory of the energy storage charge state within a future time window. As the initial condition for the recursive formula, within the time interval t=1, based on Optimize model output as well as The above formula is used to calculate the result. By analogy, we can obtain the energy storage state of charge in each time period from t=2 to t=H.

[0051] In this embodiment, the refueling action is not considered as an economic cost term in the objective function, but it participates in the solution as a decision variable in the mixed-integer linear programming joint optimization model. It forms a unified feasible region with diesel engine scheduling, energy storage regulation, and load management through inventory state recursion constraints, refueling delay constraints, refueling quantity constraints, and safety stock boundary constraints. Without directly incorporating the cost of the refueling action into the objective function, this embodiment allows the introduction of guiding and inhibiting terms into energy storage behavior, thereby enabling energy storage to not only be limited by physical constraints but also to adapt to different inventory risks.

[0052] In some embodiments, the constraints of the mixed-integer linear programming joint optimization model for each time period t are as follows: The power balance constraint satisfies the following formula: ; in, This represents the scheduled output of the diesel engine during time period t. This represents the energy storage discharge power in the t-th time period. This represents the typical load reduction power during the t-th time period. This represents the predicted renewable energy output for the t-th time period. This represents the load forecast for the t-th time period. This represents the energy storage charging power during the t-th time period.

[0053] Diesel engine operating constraints can include constraints on the dispatched output of the diesel engine and constraints on diesel engine starting variables. The dispatched output of the diesel engine in the t-th time period... The constraints are: ; in, This indicates the minimum permissible output of the diesel engine. This indicates the maximum permissible output of the diesel engine. This represents the diesel engine's start-up state variable in the t-th time period. This state variable is a 0-1 decision variable, meaning it takes the value 0 or 1.

[0054] Among the above constraints, when When, the dispatch output of the diesel engine in the t-th time period. The constraints are expressed as follows: ; This indicates that the diesel engine is stopped and its output is 0.

[0055] when When, the dispatch output of the diesel engine in the t-th time period. The constraint conditions are expressed as ; This indicates that the diesel engine is running and its output is within the allowable range.

[0056] The diesel engine start-up variable is used to characterize the start-up behavior of switching from a stopped state to a started state between adjacent time periods. The diesel engine start-up variable for the t-th time period... The constraints are: ; .

[0057] in, This represents the diesel engine start-up state variable in the t-th time period. This represents the diesel engine startup state variable during the (t-1)th time period. Specifically, it represents the initial state in this round of optimization corresponding to the current scheduling time. The actual diesel engine start-up status at the current dispatch time. .

[0058] To ensure that the startup variables strictly correspond to changes in the boot state, further settings can be made; ; ; In addition to satisfying physical boundaries, the fuel inventory trajectory within the future time window must also meet inventory safety constraints. Both hard fuel constraints and soft, reinforced equivalent inventory constraints can be retained simultaneously; therefore, fuel inventory constraints include constraints on the remaining fuel quantity and constraints on the predicted equivalent energy inventory.

[0059] Remaining fuel quantity after the t-th time period ends Represented as: ; Forecast equivalent energy inventory for the t-th time period Satisfy the following formula: ; in, Indicates minimum fuel inventory. This indicates the maximum capacity of the oil storage unit.

[0060] The above constraints mean that the safety of system operation is not determined solely by fuel inventory, but also by the ability of energy storage to replace diesel power generation under the current conditions. When fuel inventory is close to the limit but energy storage still has a high level of available energy, the system can still meet future energy security requirements through equivalent inventory; conversely, even if fuel inventory has not fallen below the hard lower limit, when the energy storage support capacity declines simultaneously, a more conservative dispatch strategy will be triggered in advance through the equivalent inventory constraint.

[0061] For example, the energy storage constraint satisfies the following formula: ; ; in, Indicates the maximum allowable charging power of the energy storage. This represents the maximum allowable discharge power of the energy storage. The first constraint is for the charging power of the energy storage, and the second constraint is for the discharging power of the energy storage.

[0062] To avoid simultaneous charging and discharging at the same time, mutually exclusive variables can be introduced. , ; .

[0063] Set refueling delay corresponding If refueling is triggered in the t-th discrete time period, the corresponding refueling amount is in the time period. The arrival and refueling delay satisfy the following formula: ; The fuel replenishment constraint satisfies the following formula: ; in, Represents the t+Lth oil The refueling time reaches the state variable within a certain time period. If A value of 1 indicates that the refueling task triggered in the t-th time period has elapsed through L. oil After a refueling delay of one time period, at time t+L oil Arriving within a certain time period; if A value of 0 indicates that no new refueling task was triggered during the t-th time period. This represents the refueling trigger variable in the t-th time period. This indicates the lower limit of a single refueling operation. This indicates the maximum amount of fuel that can be added in a single transaction. This represents the state variable indicating the arrival of refueling in the t-th time period. This represents the arrival amount corresponding to the newly triggered refueling in the t-th time period. For refueling tasks that were triggered before the current scheduling time but have not yet arrived, their arrival status is determined by the status of the tasks in transit.

[0064] Once replenishment arrives, the inventory must not exceed the maximum capacity of the storage unit, i.e. ; in, Indicates the minimum fuel inventory allowed for the oil storage unit; This indicates the maximum fuel inventory of the oil storage unit, i.e., the maximum usable capacity of the oil tank.

[0065] In scenarios where multiple batches of refueling tasks arrive in parallel, refueling tasks overlap across windows, or refueling is triggered multiple times within the same window, a refueling task index j can be introduced to represent the total amount of refueling arrivals in time period t as the sum of the arrival amounts of each refueling task in that time period, thereby expanding the modeling of multiple batches of refueling tasks in transit.

[0066] Critical loads are prioritized for protection and are not actively disconnected during normal scheduling; general loads are allowed to be reduced when necessary. Therefore, load reduction constraints satisfy the following formula: ; in, This represents the typical load reduction power during the t-th time period. This represents the upper limit of general load that can be reduced in the t-th time period. It is usually obtained by subtracting critical loads that must be guaranteed (such as radar, communication equipment, hospitals, etc.) from the total predicted load of the microgrid, or directly obtained from the microgrid's load management system (energy management system).

[0067] In this embodiment, constraints such as dynamic safety stock boundary and power balance, diesel engine operation, energy storage status, load reduction, and refueling delay are uniformly incorporated into the model's constraints, forming the basis for solving the optimization model. Since the model includes both continuous and discrete variables, such as diesel engine start-stop variables, startup variables, mutually exclusive energy storage charging / discharging variables, and power variables, the objective function and constraints can be written in linear or piecewise linear form. Therefore, the entire microgrid scheduling problem can be transformed into a mixed-integer linear programming problem and solved using a corresponding solver. By introducing SOC deviation cost and energy storage usage cost into the objective function, and introducing energy storage strategy reserves and power change limits into the constraints, fuel consumption, diesel engine start-stop, and load reduction can be optimized through the objective function and constraints. Furthermore, ineffective energy storage charging / discharging and oscillations between adjacent time periods can be actively suppressed, improving the engineering feasibility of the scheduling scheme.

[0068] In some embodiments, such as Figure 4As shown, based on the state information at the current scheduling moment, a mixed-integer linear programming joint optimization model is solved to obtain the optimal scheduling parameters at the current scheduling moment. Control is then performed on the islanded microgrid using these optimal scheduling parameters, including: S401, Based on the state information at the current scheduling moment, solve the mixed integer linear programming joint optimization model to obtain the optimal scheduling parameter sequence within the future time window, where the future time window includes H time periods and the optimal scheduling parameter sequence includes the optimal scheduling parameters for the H time periods; S402: Obtain the optimal scheduling parameters corresponding to the current scheduling time from the optimal scheduling parameter sequence, and perform control on the islanded microgrid using the optimal scheduling parameters corresponding to the current scheduling time.

[0069] Based on the objective function, constraints, and scheduling parameters (decision variables of the model) constructed above, the problem is transformed into solving a mixed-integer linear programming joint optimization model. That is, under the premise of satisfying the constraints of power balance, diesel engine operation, energy storage, fuel inventory, refueling delay, refueling quantity, and load reduction, the objective function is minimized, which means minimizing the comprehensive cost of fuel consumption cost, diesel engine start-up cost, load reduction penalty, energy storage charge state deviation cost, and energy storage usage cost. Therefore, based on the state information at the current scheduling moment, the joint control scheme that satisfies all constraints and minimizes the comprehensive cost within the future time window can be solved.

[0070] In some embodiments, obtaining the optimal scheduling parameters corresponding to the current scheduling time from the optimal scheduling parameter sequence, and performing control on the islanded microgrid using the optimal scheduling parameters corresponding to the current scheduling time, includes: Based on the optimal diesel engine start-up variables, optimal diesel engine start-up status variables, and optimal diesel engine dispatch output corresponding to the current dispatch time, start-up or stop status, start-up action, and dispatch output command are issued to the diesel engine. Based on the optimal energy storage charging power variable and the optimal energy storage charging power variable corresponding to the current scheduling time, charging power and discharging power commands are issued to the energy storage system. When the general load reduction power is greater than 0, general load reduction control is performed on the load side. If the refueling trigger variable corresponding to the current scheduling time is 1, a refueling command is triggered at the current scheduling time, and the newly triggered refueling task is added to the in-transit refueling task status. Based on the islanded microgrid after execution control, the status information of the next scheduling moment is collected.

[0071] For example, obtain the state information at the current scheduling time k. This state information is then input into the mixed-integer linear programming joint optimization model. Represented as: ; in, This indicates the current measured remaining fuel level. This indicates the current measured state of charge of the energy storage. This indicates the current measured operating status of the diesel engine. Indicates the current status of the refueling mission en route; Based on the above state information, the current scheduling time is generated. The load forecast and renewable energy output forecast for the next future time window are denoted as follows:

[0072] In the optimization model, equipment parameters and operating parameters are set, including the upper and lower limits of diesel engine output, fuel consumption model coefficients, rated energy storage capacity and charging and discharging efficiency, upper and lower limits of oil storage capacity, refueling delay, upper and lower limits of single refueling amount, and cost coefficients in the objective function.

[0073] The variables in the joint optimization model are organized into a unified optimization vector according to the time period order:

[0074] in, This represents the diesel engine dispatch output during the t-th time period. This represents the diesel engine's operating status over a time period of t. This represents the diesel engine starting variable over t time intervals. and Let these represent the energy storage charging power and energy storage discharging power in the t-th time period, respectively. This represents the typical load reduction power over a time period of t. This represents the refueling trigger variable over t time periods. This represents the state variable indicating the arrival of refueling over time intervals t. This represents the arrival amount of the newly triggered refueling task output by the joint optimization model in the t-th time period. This represents the amount of fuel remaining after the t-th time period ends. This represents the state of charge of the energy storage at the end of the t-th time period. This represents the auxiliary variable indicating the SOC deviation in the t-th time period. This represents mutually exclusive variables for energy storage charging and discharging.

[0075] in, , , , and These are integer variables. Integer variables are used in the optimization vector. The set of corresponding positions in the middle is denoted as ,Right now:

[0076] If no mutual exclusion variable for energy storage charging and discharging is set in the specific implementation... Then you can From optimization vector and set of integer variable positions Omitted in .

[0077] The established objective function and all constraints are uniformly organized into a standard mixed-integer linear programming form for computation using MATLAB solvers such as intlinprog, Gurobi, or CPLEX, expressed as:

[0078] in, The coefficient vector of the objective function. and The inequality constraint matrix and vectors are: and The equality constraint matrix and vectors are... and Optimization vectors The lower and upper bounds, For integer variables in optimizing vectors The corresponding set of positions in the middle.

[0079] In MATLAB, the `intlinprog` function can be called, and the position of the integer variable can be specified by the integer variable index vector `intcon`, thereby obtaining the current scheduling time. The optimal solution is as follows:

[0080] in, Indicates the current scheduling time The optimal optimization vector is given below. This represents the corresponding optimal objective function value.

[0081] By solving the current scheduling time Given a mixed-integer linear programming problem, we obtain the optimal sequence of scheduling parameters within the future time window, denoted as: ; The optimal scheduling parameter sequence corresponding to the t-th time period is as follows: ; There are H time periods within the future time window, therefore the optimal scheduling parameter sequence is... This includes the optimal scheduling parameters for group H.

[0082] The model also outputs the future state trajectory corresponding to the optimal scheduling parameter sequence, denoted as: .

[0083] This future state trajectory can be understood as the dynamic evolution of the remaining fuel and energy storage charge state within the islanded microgrid system in the future time window when executing according to the obtained optimal scheduling parameters.

[0084] Based on the output of the joint optimization model, the system operating state can be interpreted as follows: If the fuel replenishment trigger variable remains 0 within the future time window, and the future fuel inventory status or equivalent energy inventory status always meets the dynamic safety stock boundary constraint, then the islanded microgrid is determined to be in normal operating condition, indicating that no new fuel replenishment action needs to be triggered within the current rolling window.

[0085] If there is a situation where the refueling trigger variable is 1 within a future time window, and the corresponding refueling task forms a refueling arrival status or refueling arrival amount after the refueling delay, then the islanded microgrid is determined to be in the refueling trigger condition, indicating that the system needs to trigger the refueling task at the present or a certain time period in the future to ensure that the future inventory trajectory meets the dynamic safety stock boundary constraint.

[0086] If a general load reduction sequence greater than 0 occurs within a future time window, the isolated microgrid is determined to be in emergency power supply mode. Under conditions of significant decline in renewable energy output, persistently high load, limited oil replenishment capacity, or other extreme circumstances, the joint optimization model maintains system operation by reducing general loads and prioritizing continuous power supply to critical loads, and can output corresponding high-level power supply operation conclusions or risk warning information.

[0087] This embodiment adopts a rolling optimization control strategy based on finite prediction time domain. That is, at each current scheduling time, only the control quantity corresponding to the first time period is executed. Then, the system state is updated according to the actual running results and the rolling window is moved forward. Then, the state information is called again to solve the optimization model for the next scheduling time.

[0088] At the current scheduling time When controlling a microgrid system, only the optimal scheduling parameter sequence is executed. The optimal scheduling parameters for the first time period (i.e., the current scheduling time k) are: ; At the current scheduling moment, the following control operations are performed on the microgrid system: according to , and It issues start-up or stop status, start-up actions, and power output dispatch instructions to the diesel engine; according to and It issues charging and discharging power commands to the energy storage system; when At that time, general load reduction control is performed on the load side; when At the current scheduling time, a refueling command is triggered, and the newly triggered refueling task is added to the in-transit refueling task status. The estimated arrival time and estimated arrival quantity of this refueling task are determined by the refueling delay and the optimized refueling arrival quantity.

[0089] After the microgrid system executes the above instructions, the next scheduling time is updated based on the actual measurement results of the system. The status information. The actual remaining fuel quantity at the next scheduling moment can be directly acquired by the sensor, or it can be updated based on actual execution feedback: ; in, This indicates the actual fuel consumption of the diesel engine during the current scheduling cycle. This indicates the amount of oil actually arriving and being added to the oil storage unit during the current scheduling cycle.

[0090] The measured state of charge of the energy storage at the next scheduling moment can be directly collected by the battery management system, or it can be updated according to the actual charge and discharge response: ; in, Indicates the rated capacity of the energy storage system. and These represent the charging efficiency and discharging efficiency of the energy storage system, respectively. and These represent the actual charging power and discharging power of the energy storage system during the current scheduling cycle, respectively. This represents the time interval between two adjacent scheduling moments.

[0091] The diesel engine operating status is updated to the actual start / stop status after the current scheduling cycle has been executed: ; in, This indicates the actual start / stop status of the diesel engine after the current scheduling cycle has been executed.

[0092] Update the status of in-transit refueling tasks: If a new refueling task is triggered at the current scheduling time, add the task to the in-transit refueling task queue; subtract one scheduling cycle from the remaining arrival delay of each refueling task in the queue; if the remaining arrival delay of a refueling task decreases to zero, include the corresponding refueling amount in the current actual refueling arrival amount, and remove the task from the in-transit refueling task queue. The updated status of in-transit refueling tasks is denoted as follows. .

[0093] Therefore, the state information at the next scheduling time k+1 is: ; Based on the updated status information The load forecast sequence and renewable energy output forecast sequence within the future time window are regenerated, and the mixed integer linear programming joint optimization model for the next scheduling time is reconstructed and solved according to the above implementation method, thereby realizing rolling optimization control.

[0094] In this embodiment, the scheduling scheme for each scheduling moment is determined based on the latest inventory status, the latest load forecast, and the latest renewable energy forecast. Refueling actions, as internal decision variables of the model, are continuously revised with the rolling time window. Refueling tasks en route can be passed across time windows, ensuring the continuity and executability of refueling actions in scheduling control. Fuel inventory determines the feasible scheduling domain through a dynamic safety stock boundary, making inventory safety the central focus of scheduling. Refueling actions, diesel engine scheduling, energy storage regulation, and load management are all integrated into the same optimization model framework. Through the optimization model's predictive control scheduling update mechanism, the joint optimization results are continuously revised, thereby achieving forward-looking control of future inventory risks and executable scheduling.

[0095] Based on the same concept, such as Figure 5 As shown, this application provides an islanded microgrid optimized scheduling and refueling decision system, which includes: The data acquisition device 501 is used to acquire the status information of the islanded microgrid at any given scheduling time. The prediction module 502 is used to determine the load prediction sequence and renewable energy output prediction sequence of the islanded microgrid in the future time window, including the current scheduling time, based on the acquired state information of the islanded microgrid at the current scheduling time and historical data within the historical time window before the current scheduling time. The load gap and demand assessment module 503 is used to determine the net load gap assessment value within the future time window based on the load forecast sequence and the renewable energy output forecast sequence, and to determine the energy storage support power assessment value and the minimum diesel demand power forecast value within the future time window based on the net load gap assessment value within the future time window and the measured energy storage state of charge at the current dispatch time. Calculation module 504 is used to determine the predicted equivalent energy inventory in the future time window based on the energy storage support power assessment value and the predicted fuel inventory in the future time window, to determine the dynamic safety stock boundary in the future time window based on the minimum diesel demand power forecast value in the future time window, and to determine the reference state of charge in the future time window based on the dynamic safety stock boundary and the predicted equivalent energy inventory in the future time window. The optimization model module 505 is used to construct an objective function to characterize the comprehensive operating cost of the islanded microgrid within a future time window. A mixed-integer linear programming joint optimization model is constructed using the objective function, constraints, and scheduling parameters. The constraints include power balance constraints, diesel engine operation constraints, fuel inventory constraints, energy storage constraints, refueling delay constraints, refueling quantity constraints, and load reduction constraints. The fuel inventory constraints are determined based on the dynamic safety stock boundary and equivalent energy inventory within the future time window, and the energy storage constraints are determined based on the reference state of charge within the future time window. The control module 506 is used to solve the mixed integer linear programming joint optimization model based on the state information at the current scheduling time, obtain the optimal scheduling parameters at the current scheduling time, and perform control on the islanded microgrid with the optimal scheduling parameters.

[0096] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0097] In this invention, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of a single item or a plurality of items. For example, at least one of a, b, or c can represent: a, b, c, ab, ac, bc, or abc, where a, b, and c can be a single item or multiple items.

[0098] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0099] The various embodiments in this specification are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the apparatus embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0100] The above are merely preferred embodiments of this application and are not intended to limit the scope of protection of this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application are included within the scope of protection of this application.

Claims

1. A method for optimized scheduling and fuel replenishment decision-making in an islanded microgrid, characterized in that, The method includes: For any given scheduling time, Based on the status information of the islanded microgrid at the current scheduling time and the historical data within the historical time window before the current scheduling time, the load forecast sequence and renewable energy output forecast sequence of the islanded microgrid within the future time window including the current scheduling time are determined. Based on the load forecast sequence and the renewable energy output forecast sequence, the net load gap assessment value within the future time window is determined. Based on the net load gap assessment value within the future time window and the measured energy storage state of charge at the current dispatch time, the energy storage support power assessment value and the minimum diesel demand power forecast value within the future time window are determined. The predicted equivalent energy inventory in the future time window is determined based on the energy storage support power assessment value and the predicted fuel inventory in the future time window. The dynamic safety stock boundary in the future time window is determined based on the minimum diesel demand power forecast value in the future time window. The reference state of charge in the future time window is determined based on the dynamic safety stock boundary and the predicted equivalent energy inventory in the future time window. An objective function is constructed to characterize the comprehensive operating cost of an islanded microgrid within a future time window. A mixed-integer linear programming joint optimization model is then built using the objective function, constraints, and scheduling parameters. The constraints include power balance constraints, diesel engine operation constraints, fuel inventory constraints, energy storage constraints, refueling delay constraints, refueling quantity constraints, and load reduction constraints. Fuel inventory constraints are determined based on the dynamic safety stock boundary and equivalent energy inventory within the future time window, and energy storage constraints are determined based on the reference state of charge within the future time window. The scheduling parameters include diesel engine control parameters, energy storage control parameters, load control parameters, and refueling control parameters. Based on the state information at the current scheduling moment, the mixed integer linear programming joint optimization model is solved to obtain the optimal scheduling parameters at the current scheduling moment, and the islanded microgrid is controlled using the optimal scheduling parameters.

2. The islanded microgrid optimized scheduling and refueling decision-making method according to claim 1, characterized in that, The status information includes the current measured remaining fuel quantity, current diesel engine start / stop status, current actual diesel engine output, current energy storage system state of charge, current actual renewable energy output, current measured total load, and current status of ongoing refueling tasks. The future time window includes H time periods t, determining the energy storage support power assessment value and the minimum diesel demand power prediction value within the future time window, including: For any time period t, The net load gap assessment value for time period t is determined by the difference between the load forecast and the renewable energy output forecast for that time period t. Based on the measured state of charge of energy storage, rated capacity of energy storage, lower limit of basic allowable limit of energy storage and discharge efficiency at the current scheduling time, determine the total amount of energy that can be released by energy storage at the current scheduling time. Based on the total energy storage capacity that can be released at the current scheduling time, the maximum allowable discharge power of energy storage, and the net load gap assessment value for time period t, the energy storage support power assessment value for time period t is determined. The minimum diesel demand power forecast for time period t is determined by the difference between the energy storage support power assessment value and the net load gap. The estimated fuel consumption during the period t is determined by using the predicted minimum diesel power demand for the time period t and the start-up indication during the period t when operating at the minimum diesel power demand.

3. The islanded microgrid optimized scheduling and refueling decision-making method according to claim 2, characterized in that, Based on the energy storage support power assessment and the projected fuel inventory within the future time window, the projected equivalent energy inventory for the future time window is determined. Based on the minimum diesel demand power forecast within the future time window, the dynamic safety stock boundary for the future time window is determined. Based on the dynamic safety stock boundary and the projected equivalent energy inventory, the reference state of charge for the future time window is determined, including: For any time period t, Based on the predicted fuel inventory of the previous time period t-1, the estimated fuel consumption during time period t at the minimum diesel demand power, and the known fuel replenishment amount that has been triggered before the current scheduling time k and is expected to arrive in time period t, the predicted fuel inventory for time period t is determined using the fuel inventory recursion formula. Based on the energy storage support power assessment value for time period t, the total amount of energy that can be released from energy storage at the current scheduling time, and the equivalent coefficient that a unit of fuel can be converted into electricity, determine the energy storage equivalent fuel inventory assessment value for time period t. The estimated equivalent fuel oil inventory for the time period t and the predicted fuel oil inventory for the time period t are used to determine the predicted equivalent energy inventory for the time period t. Based on the assessed fuel consumption during the period t at the minimum diesel demand power and the safety margin term for that period t, the dynamic safety stock boundary for that period t is determined. The safety margin term for that period t is determined based on the load forecast error conversion, renewable energy forecast error conversion, critical load power, and refueling delay for that period t. Based on the dynamic safety stock boundary and the predicted equivalent energy stock for time period t, the inventory risk index for time period t is determined. Based on the inventory risk index, low-risk reference state of charge, high-risk reference state of charge, and saturation function of the risk mapping interval parameter for time period t, the reference state of charge for time period t is determined.

4. The islanded microgrid optimized scheduling and refueling decision-making method according to claim 3, characterized in that, Use the following fuel inventory recursive formula to determine the predicted fuel inventory for the t-th time period. for: ； ； Where H represents the duration of the future time window, This represents the initial conditions of the formula. This indicates the actual remaining fuel quantity at the current dispatch time. This represents the predicted fuel inventory for the previous time period t-1; This represents the estimated fuel consumption during the t-th time period when operating at the minimum diesel power requirement. This represents the known amount of fuel replenishment that has been triggered before the current scheduling time k and is expected to arrive in time period t. The remaining releasable energy from the stored energy in the t-th time period is determined using the following formula. ; ； in, This indicates the total amount of electricity that can be released from energy storage at the time of dispatch. Δt represents the energy storage support power assessment value in the t-th time period, and Δt represents the interval between two adjacent scheduling times. The energy storage equivalent fuel inventory assessment value for the t-th time period is determined using the following formula. ; ； in, This represents the equivalent energy that a unit of diesel fuel can be converted into electrical energy. The predicted equivalent energy inventory for the t-th time period is determined using the following formula. ; ； The dynamic safety stock boundary for the t-th time period Represented as; ； ； in, This indicates the lower limit of the physical safety of the fuel depot. This represents the safety margin term for the t-th time period. This represents the number of discrete time intervals corresponding to the refueling delay. This indicates the actual time length corresponding to the refueling delay. This represents the estimated fuel consumption obtained when operating at the minimum diesel power requirement during the i-th time period. , , This represents the conversion factor. This represents the adjusted load forecast error for the t-th time period. This represents the calculated error in renewable energy forecasts for the t-th time period. This represents the critical load power in the t-th time period; The following formula is used to determine the inventory risk indicator for the t-th time period. ; ； in, Represents small positive numbers; The reference state of charge in the t-th time period Represented as; ； in, Indicates a low-risk reference state of charge. Indicates a high-risk reference state of charge. and This represents the risk mapping interval parameter. This represents a saturation function.

5. The islanded microgrid optimized scheduling and refueling decision-making method according to claim 4, characterized in that, An objective function is constructed to characterize the comprehensive operating cost of an islanded microgrid within a future time window. A mixed-integer linear programming joint optimization model is then built using the objective function, constraints, and scheduling parameters, including: The diesel engine control parameters include diesel engine start-up variables, diesel engine start-up state variables, and diesel engine dispatch output; the energy storage control parameters include energy storage charging power, energy storage discharging power, and energy storage state of charge at the end of the t-th time period; the load control parameters include general load reduction power; and the refueling control parameters include refueling trigger variables, refueling arrival state variables, and the arrival amount of the new refueling task triggered by the joint optimization model at the t-th time period. The comprehensive operating costs of an isolated microgrid within a future time window include fuel consumption costs, diesel engine start-up costs, load shedding penalties, energy storage state-of-charge deviation costs, and energy storage usage costs. To minimize these comprehensive operating costs, an objective function is constructed. ; ； ； ； in, This represents the unit fuel cost coefficient. This represents the diesel engine fuel consumption in the t-th time period. This represents the dispatch output of the diesel engine in the t-th time period. This represents the diesel engine start-up status variable during the t-th time period. This represents the cost coefficient for a single start-up of a diesel engine. This represents the diesel engine starting variable during the t-th time period. This represents the general load reduction penalty factor. This represents the typical load reduction power during the t-th time period. This indicates the interval between two adjacent scheduling times. This represents the weighting coefficient for the deviation of the state of charge. This represents the energy storage state of charge deviation in the t-th time period. This represents the energy storage usage cost coefficient. This represents the energy storage charging power in the t-th time period. This represents the energy storage discharge power in the t-th time period. This represents the state of charge of the energy storage at the end of the t-th time period. This represents the fuel consumption coefficient related to diesel engine output. This indicates the basic fuel consumption coefficient for diesel engines at startup.

6. The islanded microgrid optimized scheduling and refueling decision-making method according to claim 5, characterized in that, The method further includes: Remaining fuel quantity after the t-th time period ends for; ； ； ； in, This represents the amount of fuel remaining after the (t-1)th time period. This represents the initial conditions of the formula. This represents the total amount of fuel replenished in the t-th time period. This represents the arrival amount of the newly triggered refueling task output by the joint optimization model in the t-th time period; Using the following recursive formula for the state of charge of energy storage, determine the state of charge of energy storage at the end of the t-th time period. : ； ； middle, Indicates the rated capacity of energy storage. Indicates charging efficiency. Indicates discharge efficiency. This represents the initial conditions of the formula. This indicates the measured state of charge of the energy storage at the current scheduling moment. This represents the state of charge of the energy storage at the end of the (t-1)th time period.

7. The islanded microgrid optimal scheduling and refueling decision-making method according to claim 6, characterized in that, The specific constraints are as follows: The power balance constraint satisfies the following formula: ； in, This represents the dispatched output of the diesel engine in the t-th time period. This represents the energy storage discharge power in the t-th time period. This represents the typical load reduction power during the t-th time period. This represents the predicted renewable energy output for the t-th time period. This represents the load forecast for the t-th time period. This represents the energy storage charging power during the t-th time period; The diesel engine operating constraints satisfy the following formula: ； ； in, This represents the dispatch output of the diesel engine in the t-th time period. This indicates the minimum permissible output of the diesel engine. This indicates the maximum permissible output of the diesel engine. This represents the diesel engine's start-up state variable during the t-th time period. This represents the diesel engine start-up state variable during the (t-1)th time period. This represents the diesel engine starting variable during the t-th time period; The fuel inventory constraint satisfies the following formula: ； ； in, This represents the amount of fuel remaining after the t-th time period ends. This represents the predicted equivalent energy inventory for the t-th time period. Indicates minimum fuel inventory. Indicates the maximum capacity of the oil storage unit; The energy storage constraint satisfies the following formula: ； ； in, Indicates the maximum allowable charging power of the energy storage. Indicates the maximum allowable discharge power of the energy storage. This represents the energy storage charging power in the t-th time period. This represents the energy storage discharge power during the t-th time period; The refueling delay constraint is: ； in, This represents the state variable indicating the arrival of refueling in the (t+Loil)th time period. This represents the refueling trigger variable in the t-th time period; The replenishment amount constraint satisfies the following formula: ； in, This represents the arrival amount of the newly triggered refueling task output by the joint optimization model in the t-th time period. This indicates the lower limit of a single refueling operation. This indicates the maximum amount of fuel that can be added in a single transaction. This represents the state variable indicating the arrival of refueling in the t-th time period; The load reduction constraint satisfies the following formula: ； in, This represents the typical load reduction power during the t-th time period. This represents the maximum power limit of the general load that can be reduced during the t-th time period.

8. The islanded microgrid optimized scheduling and refueling decision-making method according to claim 1, characterized in that, Based on the state information at the current scheduling moment, the mixed-integer linear programming joint optimization model is solved to obtain the optimal scheduling parameters at the current scheduling moment. Control of the islanded microgrid is then performed using these optimal scheduling parameters, including: Based on the state information at the current scheduling moment, the mixed integer linear programming joint optimization model is solved to obtain the optimal scheduling parameter sequence within the future time window, wherein the future time window includes H time periods, and the optimal scheduling parameter sequence includes the optimal scheduling parameters for the H time periods. The optimal scheduling parameters corresponding to the current scheduling time are obtained from the optimal scheduling parameter sequence, and the islanded microgrid is controlled using the optimal scheduling parameters corresponding to the current scheduling time.

9. The islanded microgrid optimal scheduling and refueling decision-making method according to claim 8, characterized in that, Obtain the optimal scheduling parameters corresponding to the current scheduling time from the optimal scheduling parameter sequence, and perform control on the islanded microgrid using the optimal scheduling parameters corresponding to the current scheduling time, including: Based on the optimal diesel engine start-up variables, optimal diesel engine start-up status variables, and optimal diesel engine dispatch output corresponding to the current dispatch time, start-up or stop status, start-up action, and dispatch output command are issued to the diesel engine. Based on the optimal energy storage charging power variable and the optimal energy storage charging power variable corresponding to the current scheduling time, charging power and discharging power commands are issued to the energy storage system. When the general load reduction power is greater than 0, general load reduction control is performed on the load side. If the refueling trigger variable corresponding to the current scheduling time is 1, a refueling command is triggered at the current scheduling time, and the newly triggered refueling task is added to the in-transit refueling task status. Based on the islanded microgrid after execution control, the status information of the next scheduling moment is collected.

10. A system for optimized scheduling and refueling decision-making in an islanded microgrid, characterized in that, The system includes: The data acquisition device is used to obtain the status information of the islanded microgrid at any given scheduling time. The prediction module is used to determine the load prediction sequence and renewable energy output prediction sequence of the islanded microgrid within a future time window, including the current scheduling time, based on the acquired state information of the islanded microgrid at the current scheduling time and historical data within the historical time window before the current scheduling time. The load gap and demand assessment module is used to determine the net load gap assessment value within the future time window based on the load forecast sequence and the renewable energy output forecast sequence, and to determine the energy storage support power assessment value and the minimum diesel demand power forecast value within the future time window based on the net load gap assessment value within the future time window and the measured energy storage state of charge at the current dispatch time. The calculation module is used to determine the predicted equivalent energy inventory in the future time window based on the energy storage support power assessment value and the predicted fuel inventory in the future time window, to determine the dynamic safety stock boundary in the future time window based on the minimum diesel demand power forecast value in the future time window, and to determine the reference state of charge in the future time window based on the dynamic safety stock boundary and the predicted equivalent energy inventory. The optimization model module is used to construct an objective function to characterize the comprehensive operating cost of the islanded microgrid within a future time window. A mixed-integer linear programming joint optimization model is constructed using the objective function, constraints, and scheduling parameters. The constraints include power balance constraints, diesel engine operation constraints, fuel inventory constraints, energy storage constraints, refueling delay constraints, refueling quantity constraints, and load reduction constraints. Fuel inventory constraints are determined based on the dynamic safety stock boundary and equivalent energy inventory within the future time window, and energy storage constraints are determined based on the reference state of charge within the future time window. The scheduling parameters include diesel engine control parameters, energy storage control parameters, load control parameters, and refueling control parameters. The control module is used to solve the mixed integer linear programming joint optimization model based on the state information at the current scheduling time, obtain the optimal scheduling parameters at the current scheduling time, and perform control on the islanded microgrid using the optimal scheduling parameters.

Images