An islanded operation control method and system based on joint optimization of energy storage and feeder

Abstract

This invention discloses an islanded operation control method and system based on joint optimization of energy storage and feeders, solving the problems of low efficiency and poor adaptability in existing islanded operation systems. The method includes: establishing an islanded power supply model and a discrete state-space description; constructing a load prediction function by combining historical load data and weather factors to obtain a future load prediction sequence; designing a multi-objective optimization cost function based on the power supply model characteristics and the prediction sequence, and setting scheduling constraints such as power boundaries, state switching, energy storage capacity, feeder connectivity, and safety limitations. Within each scheduling cycle, based on the latest prediction sequence and discrete state, a model predictive control algorithm is used to continuously solve for the power output and start / stop states of the power supply according to the cost function and constraints, and the results are sent to the control device to achieve periodic closed-loop scheduling. This method can accurately predict loads in complex environments, improve power supply reliability and system stability, extend equipment life, reduce operating costs, and achieve efficient, stable, and economical power supply.

Description

Technical Field

[0001] This invention relates to the field of distribution network emergency control and islanding operation management technology, specifically to an islanding operation control method and system based on joint optimization of energy storage and feeders. Background Technology

[0002] In the context of frequent extreme natural disasters, the islanding capability of distribution networks has become an important indicator of grid resilience. Currently, most islanding operation control strategies still rely on static switching or fixed priority methods, failing to flexibly adjust power supply schemes based on real-time load changes, energy storage status, or feeder topology. Furthermore, the islanding mode switching process is prone to voltage and frequency fluctuations, exhibiting poor stability and making it difficult to achieve continuous and stable power supply to critical loads.

[0003] The rapid response capability of energy storage devices and the structural flexibility of feeders provide a physical basis for constructing a multi-source collaborative control mechanism within an isolated system. However, there is currently a lack of a unified optimization control framework that covers the entire process of isolated system operation (modeling, scheduling, switching, and reconfiguration). Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides an islanded operation control method and system based on joint optimization of energy storage and feeders, aiming to solve the problems in the background technology.

[0005] To achieve the above objectives, the present invention provides the following technical solution: an islanded operation control method based on joint optimization of energy storage and feeder, comprising the following steps:

[0006] Step S1: Establish the islanded power source model and discrete state-space description;

[0007] Step S2: Based on historical load data and weather factors, construct a dynamic load forecast function to obtain the load forecast sequence for future time periods;

[0008] Step S3: Design a multi-objective optimization cost function based on the characteristics of the islanded power source model and the load forecast sequence;

[0009] Step S4: Set scheduling constraints, including power boundary constraints, state switching constraints, energy storage capacity constraints, and feeder connectivity and safety limits;

[0010] Step S5: Within each scheduling cycle, based on the latest load forecast sequence and discrete state, and using a model predictive control algorithm based on the multi-objective optimization cost function and scheduling constraints, the power output and start / stop state of the power source are solved in a rolling manner, and the solution results are sent to each energy storage / power source / feeder control device to achieve periodic closed-loop scheduling.

[0011] Furthermore, the specific process of step S1 is as follows:

[0012] Step S1.1: Establish a typical islanded power source model; the islanded power source includes energy storage devices, photovoltaic modules, and mobile power vehicles;

[0013] Step S1.2: Define the maximum power output, minimum power, state variables, response delay, and unit operating cost for each type of islanded power supply;

[0014] Step S1.3: Through system identification and modeling, obtain the discrete state-space form:

[0015] ;

[0016] ;

[0017] In the formula, express The state variables at any given time include energy storage capacity, power supply start / stop, and feeder load. express The state variable at any given time; express The control input vector at any given time includes power output command and feeder switching status. express The external disturbance vector at any given time, which includes load fluctuations and weather effects; express External disturbance variables at time; , , , Each represents a system model matrix, used to characterize the dynamic relationships between vectors.

[0018] Furthermore, based on historical load data and weather factors, combined with the load curve change trend, an autoregressive + exponential smoothing prediction model is used to obtain future... Load forecast sequence at each time point:

[0019] ;

[0020] In the formula, express Load forecast sequence at time point, ; express Historical / current load demand at any given moment; express Historical / current load demand at any given moment; , All of these represent the coefficients of the autoregressive term, used to characterize the autocorrelation of the load; , Both represent the coefficients of the exponential smoothing term, used to characterize the trend of the load; This represents the natural index.

[0021] Furthermore, a multi-objective optimization cost function is designed based on the characteristics of the islanded power supply model and the load forecast sequence. :

[0022] ;

[0023] In the formula, This indicates the total energy supplied by all power sources; express Load forecast sequence at any given time; It represents a small positive number with a denominator of zero; Indicates the first One power supply The start / stop status at any given moment; Indicates the first One power supply The start / stop status at any given moment; Indicates the first The unit operating cost of a power source; Indicates the first One power supply Power output at any given moment; Indicates the total number of power supplies; , , All of these represent target weighting factors.

[0024] Furthermore, the power boundary constraint is as follows: ; , They represent the first The minimum and maximum power output of each power supply;

[0025] The state transition constraint is: ;

[0026] The energy storage capacity constraint is: ; , They represent the first Upper and lower limits of the state of charge of an energy storage device; Indicates the first An energy storage device in The state of charge at any given moment;

[0027] The feeder connectivity and safety constraints are as follows: there must be no feeder backfeeding, isolated points, or overload paths.

[0028] An islanded operation control system based on joint optimization of energy storage and feeder, used to execute an islanded operation control method based on joint optimization of energy storage and feeder, comprising:

[0029] The model building module is used to establish islanded power source models and discrete state-space descriptions.

[0030] The acquisition module is used to construct a dynamic load forecasting function based on historical load data and weather factors, and to obtain the load forecast sequence for future time periods;

[0031] The function design module is used to design multi-objective optimization cost functions based on the characteristics of islanded power source models and load forecast sequences.

[0032] The constraint module is used to set scheduling constraints, including power boundary constraints, state switching constraints, energy storage capacity constraints, and feeder connectivity and safety limits.

[0033] The scheduling module is used to solve the power output and start / stop status of the power source in each scheduling cycle based on the latest load forecast sequence and discrete status, using a multi-objective optimization cost function and scheduling constraints, and employing a model predictive control algorithm. The solution results are then sent to each energy storage / power source / feeder control device to achieve periodic closed-loop scheduling.

[0034] Furthermore, in the function design module, a multi-objective optimization cost function is designed based on the characteristics of the islanded power supply model and the load forecast sequence. :

[0035] ;

[0036] In the formula, This indicates the total energy supplied by all power sources; express Load forecast sequence at any given time; It represents a small positive number with a denominator of zero; Indicates the first One power supply The start / stop status at any given moment; Indicates the first One power supply The start / stop status at any given moment; Indicates the first The unit operating cost of a power source; Indicates the first One power supply Power output at any given moment; Indicates the total number of power supplies; , , All of these represent target weighting factors.

[0037] Furthermore, in the constraint module, the power boundary constraints are as follows: ; , They represent the first The minimum and maximum power output of each power supply;

[0038] The state transition constraint is: ;

[0039] The energy storage capacity constraint is: ; , They represent the first Upper and lower limits of the state of charge of an energy storage device; Indicates the first An energy storage device in The state of charge at any given moment;

[0040] The feeder connectivity and safety constraints are as follows: there must be no feeder backfeeding, isolated points, or overload paths.

[0041] An electronic device includes a processor, a memory, and a bus, wherein the processor and the memory are connected via the bus, wherein the memory is used to store a set of program code, and the processor is used to call the program code stored in the memory to execute an islanded operation control method based on joint optimization of energy storage and feeder.

[0042] A non-volatile computer storage medium storing computer-executable instructions that execute an islanded operation control method based on joint optimization of energy storage and feeder.

[0043] Compared with existing technologies, the present invention has the following advantages:

[0044] (1) This invention constructs a unified optimization control framework covering the entire process of islanded operation modeling, scheduling, switching, and reconfiguration through the joint optimization of energy storage and feeders. This framework fully leverages the rapid response capability of energy storage devices and the structural flexibility of feeders to achieve multi-source coordinated operation of photovoltaics, energy storage, and mobile power vehicles; it not only improves energy utilization efficiency and avoids energy waste caused by the idleness or overload of a single power source, but also simplifies the scheduling operation process, reduces the need for manual intervention, and adapts to the actual needs of personnel in extreme disaster scenarios where on-site operation is inconvenient, thereby achieving efficient, stable, and economical power supply for islanded systems.

[0045] (2) This invention establishes a typical islanded power model covering energy storage devices, photovoltaic modules, and mobile power vehicles, and constructs a discrete state space description including state variables, control input vectors, and external disturbance vectors, which can accurately characterize the dynamic operation characteristics of the islanded system in real time. At the same time, by combining historical load data with weather factors such as temperature, humidity, and wind speed, an autoregressive + exponential smoothing prediction model is used to generate a load prediction sequence for future time moments, which can effectively reduce load prediction errors in complex disaster environments, provide accurate data support for islanded system scheduling, and ensure the continuous and stable power supply of key loads such as emergency command centers and temporary medical points in extreme disaster scenarios such as earthquakes, typhoons, and floods, avoiding power outages caused by load prediction deviations or inaccurate system state characterization.

[0046] (3) The multi-objective optimization cost function designed in this invention achieves balanced optimization of multiple objectives in islanded operation through the weighted coordination of supply and demand matching deviation term, equipment start-up and shutdown change term, and operation cost term: on the one hand, it ensures that the total power supply is accurately matched with the predicted load to meet the load's power demand; on the other hand, it effectively reduces the frequent start-up and shutdown operations of equipment such as energy storage and mobile power vehicles, reduces equipment mechanical wear, and extends equipment service life; at the same time, it prioritizes the scheduling of photovoltaic, energy storage and other power sources, reduces the consumption of high fuel costs of mobile power vehicles, significantly reduces the overall operating cost of the islanded system, and takes into account the reliability of emergency power supply, equipment protection and operation economy.

[0047] (4) The power boundary constraints, state switching constraints, energy storage capacity constraints, feeder connectivity and safety limitation constraints set by this invention form a comprehensive scheduling safety protection system. This system can effectively avoid equipment burnout caused by overload and inefficient operation caused by underload of each power source, prevent battery bulging and sudden life reduction of energy storage devices due to overcharging or permanent capacity loss due to over-discharging, avoid the risk of power outages such as voltage surge and line burnout caused by feeder topology faults, greatly improve the safety and reliability of islanded system operation, and reduce the system failure rate.

[0048] (5) In each scheduling cycle, this invention, based on the latest load forecast sequence and the discrete state of the islanded system, combines a multi-objective optimization cost function and scheduling constraints, and uses a model predictive control algorithm to solve the power output and start / stop status of the power source in a rolling manner. The solution results are then sent to each energy storage / power source / feeder control device to achieve periodic closed-loop scheduling. This scheduling mechanism can quickly respond to sudden disturbances during disasters, such as sudden load increases, sudden drops in photovoltaic output due to weather changes, and temporary failures of a power source. By dynamically adjusting the power allocation and feeder switching status, it maintains the voltage and frequency of the islanded system within a reasonable range, significantly enhancing the system's anti-disturbance capability and solving the problems of lag and poor adaptability in traditional static switching and fixed priority scheduling. Attached Figure Description

[0049] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0050] like Figure 1 As shown, the present invention provides a technical solution: an islanded operation control method based on joint optimization of energy storage and feeder, comprising the following steps:

[0051] Step S1: Establish the islanded power source model and discrete state-space description.

[0052] Step S1.1: Establish typical islanded power models such as energy storage devices, photovoltaic modules, and mobile power vehicles.

[0053] Step S1.2: Define the maximum power output, minimum power, state variables, response delay, and unit operating cost for each type of islanded power source.

[0054] Step S1.3: Through system identification and modeling, obtain the discrete state-space form:

[0055] ;

[0056] ;

[0057] In the formula, express State variables at any given time (energy storage capacity, power supply start / stop, feeder load, etc.). express The state variable at any given time; express Control input vector at any time (power output command, feeder switching status); express The vector of external disturbances at any given time (load fluctuations, weather effects, etc.); express External disturbance variables at any given time (such as sudden load changes, environmental changes, etc.); , , , Each represents a system model matrix, used to characterize the dynamic relationships between vectors.

[0058] Regarding the above methods (output) It should be noted that this model is a core mathematical framework built to accurately characterize the dynamic characteristics of isolated systems, comprising four main elements: state variables, control inputs, external disturbances, and output variables. It covers key statuses such as energy storage capacity, power supply start / stop, and feeder load. It serves as a control input for power commands, feeder switching, etc. Including external disturbances such as load fluctuations and weather effects, through a matrix , , , This model quantifies the dynamic relationships between variables. It breaks through the limitations of traditional static modeling, which only describes fixed states, and can reflect the state changes of isolated systems in real time due to load changes, equipment failures, etc. It provides an accurate dynamic prediction basis for subsequent MPC rolling optimization and solves the problem that the traditional model is "static and rigid" and cannot adapt to complex disaster scenarios.

[0059] Step S2: Based on historical load data and weather factors, construct a dynamic load forecast function to obtain the load forecast sequence for future times.

[0060] Based on historical load data and weather factors (temperature, humidity, wind speed), combined with the load curve trend, an autoregressive + exponential smoothing (AR-ES) prediction model is used to obtain future... Load forecast sequence at each time point:

[0061] ;

[0062] In the formula, express Load forecast sequence at time point, ; express Historical / current load demand at any given moment; express Historical / current load demand at any given moment; , All of these represent the coefficients of the autoregressive term, used to characterize the autocorrelation of the load; , Both represent the coefficients of the exponential smoothing term, used to characterize the trend of the load; This represents the natural index.

[0063] Regarding the dynamic load forecasting function in the above method It should be noted that this function is built on an autoregressive-exponential smoothing (AR-ES) model, which integrates historical load data with meteorological factors such as temperature, humidity, and wind speed to generate future load data. This function generates a load forecast sequence for each time point. Its design logic lies in the fact that isolated loads are significantly affected by the post-disaster environment (such as the deployment of temporary relief equipment and power fluctuations caused by weather changes), and traditional forecasting methods relying solely on historical data have large errors. This function improves the timeliness and accuracy of load forecasting by introducing meteorological factors to correct forecast biases. It provides a reliable load benchmark for Model Predictive Control (MPC) algorithm-based optimized scheduling, avoiding excessive energy storage discharge or insufficient power supply to the load due to inaccurate forecasts, and solving the problem of the disconnect between traditional load forecasting and actual load during disasters.

[0064] Step S3: Design a multi-objective optimization cost function based on the characteristics of the islanded power source model and the load forecast sequence. :

[0065] ;

[0066] In the formula, This indicates the total energy supplied by all power sources; express Load forecast sequence at any given time; It represents a small positive number with a denominator of zero; Indicates the first One power supply The start / stop status at any given moment; Indicates the first One power supply The start / stop status at any given moment; Indicates the first The unit operating cost of a power source; Indicates the first One power supply Power output at any given moment; Indicates the total number of power supplies; , , All of these represent target weighting factors.

[0067] Multi-objective optimization cost function Includes three items:

[0068] Supply and demand mismatch item It is used to measure the degree of deviation between the total energy supply of the system and the predicted load.

[0069] Equipment start-up and shutdown changes Frequent start-stop operations are penalized to extend equipment life and improve operational stability.

[0070] Operating cost item This represents the cost per unit of power multiplied by the power output, constraining the economic objective.

[0071] The innovation lies in:

[0072] Using a supply-demand deviation normalization function instead of the traditional absolute error or squared error enhances the adaptability to loads of different scales.

[0073] Using "state differences" to characterize start-stop frequency, rather than introducing complex penalty factors, makes the calculation simpler.

[0074] The three elements together form a multi-objective optimization framework that balances security, stability, and economy.

[0075] Compared to the single objective (such as only ensuring power supply) or static strategy in traditional islanded control, this function takes into account power supply reliability, equipment stability and economy through multi-objective weighted coordination, which solves the contradiction of "paying for one thing but losing another" in traditional methods, and enables the islanded system to still operate efficiently when the load fluctuates and the energy storage status changes.

[0076] Step S4: Set scheduling constraints, including power boundary constraints, state switching constraints, energy storage capacity constraints, and feeder connectivity and safety limits.

[0077] The power boundary constraints are: ; , They represent the first The minimum and maximum power output of each power supply.

[0078] The state transition constraint is: .

[0079] Energy storage capacity constraints are: ; , They represent the first Upper and lower limits of the state of charge of an energy storage device; Indicates the first An energy storage device in The state of charge at any given moment.

[0080] The feeder connectivity and safety constraints are as follows: there must be no feeder backfeeding, isolated points, or overload paths.

[0081] Step S5: Within each scheduling cycle, based on the latest load forecast sequence and discrete state, and using the multi-objective optimization cost function and scheduling constraints, the model predictive control algorithm (MPC) is employed to solve the power output and start / stop state of the power source on a rolling basis, and the solution results are sent to each energy storage / power source / feeder control device to achieve periodic closed-loop scheduling.

[0082] An islanded operation control system based on joint optimization of energy storage and feeder, used to execute an islanded operation control method based on joint optimization of energy storage and feeder, comprising:

[0083] The model building module is used to establish islanded power source models and discrete state-space descriptions.

[0084] The acquisition module is used to construct a dynamic load forecasting function based on historical load data and weather factors, and to obtain the load forecast sequence for future time periods;

[0085] The function design module is used to design multi-objective optimization cost functions based on the characteristics of islanded power source models and load forecast sequences.

[0086] The constraint module is used to set scheduling constraints, including power boundary constraints, state switching constraints, energy storage capacity constraints, and feeder connectivity and safety limits.

[0087] The scheduling module is used to solve the power output and start / stop status of the power source in each scheduling cycle based on the latest load forecast sequence and discrete status, using a multi-objective optimization cost function and scheduling constraints, and employing a model predictive control algorithm. The solution results are then sent to each energy storage / power source / feeder control device to achieve periodic closed-loop scheduling.

[0088] Furthermore, in the function design module, a multi-objective optimization cost function is designed based on the characteristics of the islanded power supply model and the load forecast sequence. :

[0089] ;

[0090] In the formula, This indicates the total energy supplied by all power sources; express Load forecast sequence at any given time; It represents a small positive number with a denominator of zero; Indicates the first One power supply The start / stop status at any given moment; Indicates the first One power supply The start / stop status at any given moment; Indicates the first The unit operating cost of a power source; Indicates the first One power supply Power output at any given moment; Indicates the total number of power supplies; , , All of these represent target weighting factors.

[0091] Furthermore, in the constraint module, the power boundary constraints are as follows: ; , They represent the first The minimum and maximum power output of each power supply;

[0092] The state transition constraint is: ;

[0093] The energy storage capacity constraint is: ; , They represent the first Upper and lower limits of the state of charge of an energy storage device; Indicates the first An energy storage device in The state of charge at any given moment;

[0094] The feeder connectivity and safety constraints are as follows: there must be no feeder backfeeding, isolated points, or overload paths.

[0095] An electronic device includes a processor, a memory, and a bus, wherein the processor and the memory are connected via the bus, wherein the memory is used to store a set of program code, and the processor is used to call the program code stored in the memory to execute an islanded operation control method based on joint optimization of energy storage and feeder.

[0096] A non-volatile computer storage medium storing computer-executable instructions that execute an islanded operation control method based on joint optimization of energy storage and feeder.

[0097] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An islanded operation control method based on joint optimization of energy storage and feeder, characterized in that, Includes the following steps: Step S1: Establish the islanded power source model and discrete state-space description; Step S2: Based on historical load data and weather factors, construct a dynamic load forecast function to obtain the load forecast sequence for future time periods; Step S3: Design a multi-objective optimization cost function based on the characteristics of the islanded power source model and the load forecast sequence; Step S4: Set scheduling constraints, including power boundary constraints, state switching constraints, energy storage capacity constraints, and feeder connectivity and safety limits; Step S5: Within each scheduling cycle, based on the latest load forecast sequence and discrete state, and using the multi-objective optimization cost function and scheduling constraints, the model predictive control algorithm is adopted to solve the power output and start / stop state of the power source in a rolling manner, and the solution results are sent to each energy storage / power source / feeder control device to achieve periodic closed-loop scheduling. The specific process of step S1 is as follows: Step S1.1: Establish a typical islanded power source model; the islanded power source includes energy storage devices, photovoltaic modules, and mobile power vehicles; Step S1.2: Define the maximum power output, minimum power, state variables, response delay, and unit operating cost for each type of islanded power supply; Step S1.3: Through system identification and modeling, obtain the discrete state-space form: ； ； In the formula, express The state variables at any given time include energy storage capacity, power supply start / stop, and feeder load. express The state variable at any given time; express The control input vector at any given time includes power output command and feeder switching status. express The external disturbance vector at any given time, which includes load fluctuations and weather effects; express External disturbance variables at time; , , , Each represents a system model matrix, used to characterize the dynamic relationships between vectors; Based on historical load data and weather factors, combined with the trend of load curve changes, an autoregressive + exponential smoothing prediction model is used to obtain future... Load forecast sequence at each time point: ； In the formula, express Load forecast sequence at time point, ; express Historical / current load demand at any given moment; express Historical / current load demand at any given moment; , All of these represent the coefficients of the autoregressive term, used to characterize the autocorrelation of the load; , Both represent the coefficients of the exponential smoothing term, used to characterize the trend of the load; Indicates the natural index; Design a multi-objective optimization cost function based on the characteristics of islanded power supply models and load forecast sequences. : ； In the formula, This indicates the total energy supplied by all power sources; express Load forecast sequence at any given time; Represents a positive number whose denominator is zero; Indicates the first One power supply The start / stop status at any given moment; Indicates the first One power supply The start / stop status at any given moment; Indicates the first The unit operating cost of a power source; Indicates the first One power supply Power output at any given moment; Indicates the total number of power supplies; , , All of these represent target weighting factors.

2. The islanding operation control method based on joint optimization of energy storage and feeder as described in claim 1, characterized in that: The power boundary constraint is: ; , They represent the first The minimum and maximum power output of each power supply; The state transition constraint is: ; The energy storage capacity constraint is: ; , They represent the first Upper and lower limits of the state of charge of an energy storage device; Indicates the first An energy storage device in The state of charge at any given moment; The feeder connectivity and safety constraints are as follows: there must be no feeder backfeeding, isolated points, or overload paths.

3. An islanded operation control system based on joint optimization of energy storage and feeder, used to execute the islanded operation control method based on joint optimization of energy storage and feeder as described in any one of claims 1-2, characterized in that, include: The model building module is used to establish islanded power source models and discrete state-space descriptions. The acquisition module is used to construct a dynamic load forecasting function based on historical load data and weather factors, and to obtain the load forecast sequence for future time periods; The function design module is used to design multi-objective optimization cost functions based on the characteristics of islanded power source models and load forecast sequences. The constraint module is used to set scheduling constraints, including power boundary constraints, state switching constraints, energy storage capacity constraints, and feeder connectivity and safety limits. The scheduling module is used to solve the power output and start / stop status of the power source in each scheduling cycle based on the latest load forecast sequence and discrete status, using a multi-objective optimization cost function and scheduling constraints, and employing a model predictive control algorithm. The solution results are then sent to each energy storage / power source / feeder control device to achieve periodic closed-loop scheduling.

4. An electronic device, characterized in that, The system includes a processor, a memory, and a bus. The processor and the memory are connected via the bus. The memory is used to store a set of program codes, and the processor is used to call the program codes stored in the memory to execute the islanded operation control method based on joint optimization of energy storage and feeder as described in any one of claims 1-2.

5. A non-volatile computer storage medium storing computer-executable instructions, characterized in that, The computer can execute instructions to perform an islanded operation control method based on joint optimization of energy storage and feeder, as described in any one of claims 1-2.

Images