Campus microgrid group collaborative optimization method and system considering load space-time asynchronous response

Abstract

A collaborative optimization method and system for a cluster of microgrids in industrial parks, considering the asynchronous spatiotemporal response of loads, includes: collecting operational data from each microgrid in the industrial parks; classifying the loads in the industrial parks into conventional loads and adjustable loads according to load response characteristics, and constructing a response time-delay stochastic model for the adjustable loads; the response time-delay stochastic model discretizes the stochastic response time delay of the adjustable loads into multiple typical response time delays; based on the response time-delay stochastic model of the adjustable loads, and combined with the operational data of each microgrid in the industrial parks, constructing a low-carbon economic dispatch optimization model for the multi-industry microgrid clusters with the objective of minimizing overall operating costs, wherein the dispatch optimization model includes operational constraints for adjustable loads that take into account the stochastic response time delay; and solving the dispatch optimization model to obtain a collaborative dispatch operation scheme for the multi-industry clusters. This invention fully considers factors such as information transmission delays that exist in the actual response process of adjustable loads, and can realize low-carbon economic collaborative operation of multi-industry microgrid clusters.

Description

Technical Field

[0001] This invention belongs to the field of integrated energy system and microgrid coordinated dispatch technology in industrial parks, and in particular relates to a collaborative optimization method and system for industrial park microgrid groups that considers the asynchronous response of load in time and space. Background Technology

[0002] With the widespread integration of renewable energy sources such as wind and solar power into industrial parks, park microgrids have become an important means to improve the absorption of new energy and overall energy efficiency. In existing technologies, park microgrids typically include photovoltaic power generation units, wind turbine generators, energy storage systems, gas turbines, and various electrical loads. The dispatch center optimizes the scheduling of purchased and sold power, energy storage charging and discharging power, and conventional unit output based on load demand and distributed power generation forecasts to achieve economical or low-carbon operation.

[0003] Existing technologies address both demand response modeling, source-load-storage coordination control, and low-carbon optimized operation for individual microgrids, and are also beginning to study collaborative scheduling issues such as energy interaction between multiple industrial parks and multiple microgrids, shared energy storage, and coordinated electricity-carbon trading. These methods typically acquire load and wind / solar output forecast data for each industrial park, establish an optimized scheduling model, and solve for operational strategies for each time period by incorporating demand response constraints.

[0004] However, most existing technologies assume that adjustable loads can respond instantly and accurately after receiving dispatch instructions, or only use idealized time windows to describe their regulation characteristics. They fail to fully consider factors such as information transmission delays, equipment startup preparation time, and user behavioral inertia during the actual response process. They also fail to adequately consider the spatial asynchronous response issues caused by differences in user types, equipment conditions, and energy consumption habits between different zones. This can easily lead to power balance deviations, inaccurate energy storage dispatch, insufficient renewable energy absorption, and reduced mutual assistance between zones, ultimately affecting the economic efficiency and low-carbon operation of multi-zone microgrid clusters.

[0005] Chinese patent application CN121602389A discloses a stochastic optimization scheduling method for photovoltaic-storage-DC-flexible microgrids that considers demand response delay. This method introduces the delay response characteristics of three types of flexible loads—shiftable, transferable, and reduceable—when constructing the microgrid optimization scheduling model. However, it still has the following shortcomings: This method mainly addresses the delay response problem of flexible loads within a single microgrid, focusing on describing the time lag between receiving scheduling instructions and actual execution of adjustable loads. It does not fully consider the spatial asynchronous response differences arising from variations in load composition, equipment operating characteristics, user behavior, and response capabilities across different zones. Furthermore, this method fails to coordinate the actual response states of adjustable loads in different zones with power interaction, energy sharing, and the overall power balance of the microgrid group. This makes it difficult to uniformly coordinate energy storage systems, conventional generating units, and purchased and sold power based on the actual response deviations of each zone. When applied to multi-zone microgrid groups, it easily leads to inconsistencies between planned and actual response power, resulting in fluctuations in inter-zone power exchange, inaccurate scheduling of energy storage and conventional generating units, reduced renewable energy absorption capacity, and impacting the economic efficiency and low-carbon nature of the microgrid group operation. Furthermore, this method does not comprehensively consider inter-park collaboration and carbon emission trading mechanisms, making it difficult to meet the needs of low-carbon economic collaborative operation of multi-park microgrid clusters.

[0006] Therefore, it is necessary to propose a low-carbon economic dispatch method for industrial park microgrid groups that takes into account the spatiotemporal asynchronous response of adjustable loads. Summary of the Invention

[0007] To address the shortcomings of existing technologies, this invention provides a collaborative optimization method and system for industrial microgrid groups that considers the asynchronous spatiotemporal response of loads.

[0008] The present invention adopts the following technical solution.

[0009] In a first aspect, this invention discloses a collaborative optimization method for a cluster of microgrids in a park that considers the asynchronous spatiotemporal response of loads, the method comprising the following steps: Step 1: Collect operational data of the microgrids in each park; Step 2: Based on the load response characteristics, the park load is divided into conventional load and adjustable load, and a response time delay stochastic model of the adjustable load is constructed; the response time delay stochastic model discretizes the random response time delay of the adjustable load into multiple typical response time delays; Step 3: Based on the response time-delay stochastic model of the adjustable load, and combined with the operation data of each park microgrid, construct a low-carbon economic dispatch optimization model for multi-park microgrid groups with the goal of minimizing the overall operating cost. The dispatch optimization model includes adjustable load operation constraints that take into account the response time-delay stochasticity. Step 4: Solve the scheduling optimization model to obtain a multi-park collaborative scheduling operation scheme.

[0010] More preferably, In step 2, the adjustable load includes time-shifted load, transferable load, and load that can be reduced; the response time-delay stochastic model of the adjustable load is specifically as follows:

[0011]

[0012] in, The random response time delay for adjustable loads; The time delay discretization of the random response is the first... k A typical value represents the park Inner The first type of adjustable load A typical response time delay; Indicates the park Inner The adjustable load class has a time delay of 100° in random response. The probability, The corresponding probability value; K This represents the total number of typical response time delays.

[0013] More preferably, The probability value Specifically, the following methods are used to determine the load response time delay: Historical control and operation data of the adjustable load in the industrial park are collected; load response time delay samples are statistically analyzed, whereby the load response time delay samples are the duration data between the issuance of the control command and the actual load response time; the load response time delay interval is determined based on the sample distribution range; and the total number of typical response time delays is selected according to the daily dispatch calculation accuracy requirements of the power grid. K The load response time delay interval is divided into K Each sub-interval is used as a characteristic time point to represent all time-delay conditions within its sub-interval. The frequency of the load response time delay sample falling into each sub-interval is statistically analyzed, and the corresponding proportion is obtained after normalization. This proportion is used as the probability value of the typical response time delay value within the corresponding sub-interval. .

[0014] More preferably, In step 3, the adjustable load operation constraints that take into account the randomness of response time delay include: load shifting operation constraints that take into account the randomness of response time delay for time-shifted loads, power transfer constraints that take into account the randomness of response time delay for transferable loads, and power reduction constraints that take into account the randomness of response time delay for loads that can be reduced.

[0015] More preferably, The load shifting operation constraints corresponding to the time-shifted load, which take into account the randomness of response time delay, specifically include:

[0016]

[0017]

[0018]

[0019]

[0020] in, For the park Middle, No In a typical response time delay scenario, time-shifted load... Adjusting power at any time; For the park middle Original power before time-shifted load optimization; For the park middle Total regulating power after time-shifted load optimization; K This represents the total number of typical response time delays; For the first The probability value corresponding to a typical response time delay; , These are Boolean variables, representing t, , and respectively. m The time-shifted load is in a power shifting state at any given moment. A value of 1 indicates that it is in a power shifting state, and a value of 0 indicates that it is not in a power shifting state. This is the minimum continuous operating time required once the time-shifted load is started; For the park i Time-shifted load in the first Response delay in a typical response delay scenario; It is a Boolean variable, representing the first... k In a typical response time delay scenario, the time-shifted load at time... The actual response trigger flag is set to 1 if the response is triggered and 0 if the response is not triggered. This is a Boolean variable representing the time-shifted load. t The control command issuance flag is set to 1 if issued and 0 if not issued. , These are the time windows during which power shifting is permitted for time-shifted loads. The lower limit time and the upper limit time.

[0021] More preferably, The power transfer constraints corresponding to the transferable load, taking into account the randomness of the response time delay, specifically include:

[0022]

[0023]

[0024]

[0025]

[0026]

[0027] in, For the park middle Total regulating power optimized for load transfer at any time; For the park middle The original power before load transfer optimization; , These are the time windows during which power transfer is permitted for transferable loads. The lower limit time and the upper limit time; For the first The probability value corresponding to a typical response time delay; For the park Middle, No In a typical response time delay scenario, the transferable load is Adjusting power at any time; , The transferable loads are respectively The upper and lower limits of power at any given time; , These are Boolean variables, representing t, , and respectively. m This indicates whether the transferable load is in operation at any given time; a value of 1 indicates that it is in operation, and a value of 0 indicates that it is not in operation. This is a Boolean variable, representing the transferable load. t The control command issuance flag is set to 1 if issued and 0 if not issued. The shortest duration of the transferable load; For the park i Transferable load in the first Response delay in a typical response delay scenario; It is a Boolean variable, representing the first... k In a typical response time delay scenario, the transferable load at time... The actual response trigger flag is set to 1 if the response is triggered and 0 if the response is not triggered.

[0028] More preferably, The power reduction constraints that take into account the randomness of response time delays corresponding to the load reduction capacity specifically include:

[0029]

[0030]

[0031]

[0032]

[0033]

[0034] in, For the park middle Load power after time reduction; For the park Middle, No In a typical response time delay scenario Reduce the load power at any time; For the park middle Load power before the moment of reduction; For the first The probability value corresponding to a typical response time delay; For the park middle The load reduction rate at any given time; , These are Boolean variables, representing respectively t , m This indicates whether the load reduction is in a state of operation at any given time; a value of 1 indicates that the load is in a state of operation ... load reduction; This represents the maximum load reduction rate. For the park i Load can be reduced in the first Response delay in a typical response delay scenario; It is a Boolean variable, representing the first... k In a typical response time delay scenario, the load can be reduced at time [time]. The actual response trigger flag is set to 1 if the response is triggered and 0 if the response is not triggered. This is a Boolean variable, representing the load that can be reduced. t The control command issuance flag is set to 1 if issued and 0 if not issued. , These are the time windows during which power reduction can be implemented, allowing for load reduction. The lower limit time and the upper limit time; This is the lower limit for the duration during which load reduction is possible.

[0035] Secondly, this invention discloses a collaborative optimization system for a microgrid cluster in a park that considers the asynchronous spatiotemporal response of load based on the aforementioned method, including a data acquisition module, a response time-delay stochastic model construction module for adjustable loads, a low-carbon economic dispatch optimization model construction module for a multi-park microgrid cluster, and a dispatch optimization model solving module. The data acquisition module collects operational data from the microgrids in each park. The module for constructing a response time-delay stochastic model for adjustable loads divides the park load into conventional loads and adjustable loads according to the load response characteristics, and constructs a response time-delay stochastic model for adjustable loads; the response time-delay stochastic model discretizes the random response time delay of adjustable loads into multiple typical response time delays; The module for constructing a low-carbon economic dispatch optimization model for multi-park microgrid groups, based on the response time-delay stochastic model of the adjustable load and combined with the operation data of each park microgrid, constructs a low-carbon economic dispatch optimization model for multi-park microgrid groups with the goal of minimizing the overall operating cost. The dispatch optimization model includes adjustable load operation constraints that take into account the response time-delay stochasticity. The scheduling optimization model solving module solves the scheduling optimization model to obtain a multi-park collaborative scheduling operation scheme.

[0036] Thirdly, the present invention provides a terminal, including a processor and a storage medium; The storage medium is used to store instructions; The processor is configured to operate according to the instructions to perform the steps of the method according to any one of the first aspects of the present invention.

[0037] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described in any one of the first aspects of the present invention.

[0038] The beneficial effects of this invention are compared with those of the prior art: This invention, by taking into account the time lag between planned and actual responses of adjustable loads and the response differences between different industrial parks, makes the established scheduling model more consistent with the actual operating characteristics of multi-park microgrid groups, thereby improving the executability and engineering applicability of the scheduling results. Simultaneously, by coordinating and optimizing renewable energy output, energy storage system charging and discharging, gas turbine output, energy sharing between industrial parks, and power purchase and sale behavior with the main grid, it can effectively improve the collaborative operation capability and overall power balance level among multiple industrial parks. Furthermore, by introducing a load spatiotemporal asynchronous response model, the compensation process of energy storage systems and conventional units can be arranged more accurately, improving renewable energy absorption capacity and reducing the risk of wind and solar curtailment. In conjunction with a carbon emission trading mechanism, this invention can also balance the economic efficiency and low-carbon operation of the system. Attached Figure Description

[0039] Figure 1 This is a flowchart illustrating a collaborative optimization method for a microgrid group in a park that considers the asynchronous spatiotemporal response of load according to the present invention. Figure 2 This is the load-sunlight curve in Embodiment 2 of the present invention; Figure 3 This is the power balance diagram of each park in Scheme 1 of Embodiment 2 of the present invention; Figure 4 This refers to the power interaction between different campuses in Scheme 1 of Embodiment 2 of the present invention; Figure 5 This refers to the actual carbon emissions of each scheme in Embodiment 2 of the present invention; Figure 6 This is the power balance diagram of each park in Scheme 2 of Embodiment 2 of the present invention; Figure 7 This refers to the power interaction between different campuses in Scheme 2 of Embodiment 2 of the present invention. Detailed Implementation

[0040] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of this invention. The embodiments described in this application are merely some embodiments of this invention, and not all embodiments. Based on the spirit of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the protection scope of this invention.

[0041] Example 1: like Figure 1As shown, this invention discloses a collaborative optimization method for a microgrid cluster in a park that considers the asynchronous spatiotemporal response of loads. The method is applied to a microgrid cluster composed of multiple park microgrids. For each park, a collaborative scheduling model is established that includes wind power, photovoltaic, energy storage system, gas turbine, conventional load and adjustable load. At both the park level and the cluster level, the power interaction between parks, the power purchase and sale behavior with the main grid and the carbon emission trading mechanism are considered simultaneously to form a low-carbon economic optimization scheduling framework.

[0042] A single campus microgrid includes: 1) Distributed renewable energy units, including photovoltaic power generation units and / or wind turbine generators; 2) Energy storage system, used to perform charging, discharging and state of charge regulation; 3) Conventional units, preferably gas turbines; 4) Load units, including non-adjustable conventional loads and adjustable loads; 5) The dispatch center is used to receive load, wind power, and photovoltaic forecast information from each park and generate dispatch instructions for each time period; 6) Inter-park interconnection lines and power purchase / sale interfaces connected to the main power grid; 7) Carbon trading settlement module, used to calculate carbon emissions and settle transaction costs based on electricity purchase, generator power generation and renewable energy consumption.

[0043] Furthermore, the method includes the following steps: Step 1: Collect operational data of the microgrids in each park; The operational data includes, but is not limited to, load power, wind power forecast output, photovoltaic power forecast output, energy storage parameters, gas turbine parameters, electricity purchase and sale price, carbon emission factor, and adjustable load parameters.

[0044] Step 2: Based on the load response characteristics, the park load is divided into conventional load and adjustable load, and a time-delay stochastic model of the adjustable load response is constructed; The so-called regular loads are electrical loads whose power cannot be actively adjusted during the scheduling cycle and whose operating time is fixed, and they do not participate in scheduling optimization; The adjustable load includes time-shifted load, transferable load, and load that can be reduced; The time-shifted load refers to a load whose overall operation cannot be interrupted, but whose start-up time can be shifted entirely within an allowable time window, such as washing machines and dishwashers; The transferable load refers to a load whose electricity consumption remains constant within the scheduling cycle, but whose operating time can be changed within the permitted period, such as electric vehicle charging load; The term "reducible load" refers to a load whose power can be temporarily reduced, such as air conditioning, while meeting basic energy needs.

[0045] For the aforementioned adjustable loads, instead of assuming immediate execution upon receiving the instruction, a response time delay random variable is introduced to characterize the delay from "receiving the scheduling instruction" to "actually starting execution." The response time delay is a discrete random variable, taking values ​​of one or more preset time delay levels, each level corresponding to a probability distribution. These are described in this way as follows: 1) Asynchronicity in the time dimension, that is, there is a lag in load response within the same park; 2) Asynchronicity in spatial dimension, that is, the load response patterns and response ratios differ in different parks.

[0046] For the The park is in the first The total load demand for each scheduling period Represented as:

[0047] In the formula: for The normal load at any given time; , , The park middle Time-shifted load, transferable load, and load that can be reduced at any given moment.

[0048] Various adjustable loads are represented in the scheduling model through two types of variables: planned response power and actual response power.

[0049] Define the load type set as ,in , , These correspond to time-shifted loads, transferable loads, and loads that can be reduced, respectively.

[0050] For the park Load type The response time delay uses a random variable. express; Since load response delay is a continuous random variable, directly embedding it into the scheduling optimization model would significantly increase the difficulty of solving the model and reduce the timeliness of intraday scheduling. Therefore, this invention adopts an interval discretization strategy, using discrete points to equivalently represent all delay samples within the corresponding interval. This reduces the solution complexity of the stochastic scheduling model while ensuring the accuracy of scheduling calculations, thus balancing engineering practicality and computational efficiency. Furthermore, the random variable Discretization processing, discrete values , indicating the park Inner The first type of adjustable load A typical response time delay; among which... This represents the total number of typical response time delays.

[0051] The time-delay stochastic model for the adjustable load is specifically as follows:

[0052]

[0053] in, The random response time delay for adjustable loads; The time delay discretization of the random response is the first... k A typical value represents the park Inner The first type of adjustable load A typical response time delay; Indicates the park Inner The adjustable load class has a time delay of 100° in random response. The probability, The corresponding probability value; K This represents the total number of typical response time delays.

[0054] Furthermore, the probability value Determine as follows: Historical control and operation data of the park's adjustable load are collected, and load response time delay samples are statistically analyzed. These load response time delay samples are the duration data between the issuance of the control command and the actual load response time. The load response time delay interval is determined based on the sample distribution range. The total number of typical response time delays is selected based on the accuracy requirements of the power grid's daily dispatch calculations. K The load response time delay interval is divided into K The load response time delay samples are divided into several sub-intervals, and the characteristic time points of each sub-interval are taken as typical values ​​of the response time delay. The frequency of the load response time delay samples falling into each sub-interval is statistically analyzed, and the corresponding proportion of each sub-interval is obtained after normalization. This proportion is used as the probability value corresponding to the typical value of the response time delay within the sub-interval. ; , These are the minimum and maximum time delays corresponding to the load response time delay interval, respectively. Preferably, Take the 5th percentile of all effective load response time delay samples. Take the 95th percentile of all effective load response time delay samples; Preferably, the characteristic time point of each sub-interval is the right endpoint of each sub-interval.

[0055] To facilitate understanding of the above probability values Based on the calculation process and the actual operating characteristics of the adjustable load in the park, the following exemplary implementation method is provided: Historical control operation data of this type of adjustable load within the target park over the past six months were collected, and time delay samples from load receiving control commands to power response were extracted. Based on the distribution range of all time delay samples, the maximum load response time delay interval was determined to be [0, 1] h. Considering the grid's daily dispatch step size of 0.25 h and the required calculation accuracy, the number of discrete time periods was selected as 4, dividing the maximum response time delay interval into 4 sub-intervals: [0, 0.25] h, (0.25, 0.5] h, (0.5, 0.75] h, and (0.5, 0.75] h. h, (0.75, 1]h, select the right endpoint of each interval as the characteristic time point, and use the characteristic time point to represent all time delay conditions in the corresponding sub-interval, and obtain the discrete time delay representative values ​​as 0.25, 0.5, 0.75, 1 respectively; further, count the frequency of historical time delay samples falling into each sub-interval, normalize the frequency to obtain the proportion of each time delay level executing the control command, which is the probability of the response time delay taking the corresponding discrete value, which are 0.16, 0.34, 0.36 and 0.14 respectively.

[0056] The parameters above are only illustrative settings and can be adjusted flexibly according to the actual scenario.

[0057] Step 3: Based on the response time-delay stochastic model of the adjustable load, and combined with the operation data of each park microgrid, construct a low-carbon economic dispatch optimization model for multi-park microgrid groups with the goal of minimizing the overall operating cost. The dispatch optimization model includes adjustable load operation constraints that take into account the response time-delay stochasticity. The overall operating cost includes at least: 1) External power grid purchase costs and electricity sales revenue; 2) Operating costs of the energy storage system; 3) Cost of gas turbine power generation; 4) Adjustable load compensation costs; 5) Carbon emission trading costs.

[0058] The carbon emission trading cost is calculated based on the system's actual carbon emissions, carbon emission quotas, and carbon emission reduction benefits obtained from converting renewable energy power generation into green certificates.

[0059] The objective function is expressed as minimizing the overall operating cost of all parks within the scheduling period, specifically:

[0060]

[0061]

[0062]

[0063]

[0064] In the formula: , , , , They are respectively Time Park The costs of purchasing electricity from the external power grid, operating costs of energy storage systems, gas turbine power generation costs, carbon emission trading costs, and adjustable load compensation costs; , They are respectively Time Park The power output purchased and sold to external power grids; , They are respectively Time Park External grid purchase and sale price of electricity; For the park The unit operating cost of medium-sized energy storage systems; , They are respectively Time Park The charging and discharging power of the medium-capacity energy storage system; , These refer to the charging and discharging efficiency of the energy storage system; For the park gas turbines in Output power at any moment , and These are the coefficients for various aspects of the cost of gas turbine power generation; For the park gas turbines in Start / stop status indicators at any time; This refers to actual carbon emissions; For carbon emission quotas; Carbon emission allowances obtained through converting renewable energy into green certificates; The unit price for trading carbon emission rights.

[0065] The total carbon emissions, carbon emission allowances, and carbon emission allowances obtained from renewable energy conversion green certificates are calculated as follows:

[0066] In the formula: for Carbon emissions generated by the power purchased at any given time; Carbon emissions per unit power of a gas turbine; Carbon emissions per unit of power; , These are the conversion coefficient and the quantitative coefficient for converting renewable energy power generation into green certificates, respectively. For the park middle Solar power output at all times For the park middle The output of the wind turbine at any given moment.

[0067] Various adjustable load compensation costs specifically include: Total adjustable load unit cost:

[0068] Time-shifted load compensation cost:

[0069] Cost of transferable load compensation:

[0070] This can reduce load compensation costs:

[0071] In the formula: To compensate for costs associated with adjustable load; , , These are the total compensation costs for time-shifted loads, transferred loads, and loads that can be reduced, respectively. , , It is the subsidy coefficient for its corresponding unit power; , The park middle Real-time load shifting regulation power; transferable load regulation power; Adjustable power based on load reduction; For the park middle The load reduction rate at any given time.

[0072] The scheduling model includes the following constraints: 1) Power balance constraints Establish a power balance relationship for each park during each scheduling period to maintain a balance between renewable energy output, gas turbine output, energy storage charging and discharging power, external grid power purchase and sale, inter-park power exchange, and total load demand.

[0073] The corresponding power balance mathematical model is as follows:

[0074] In the formula, For the park middle Total load demand at any given time; microgrid Flow to microgrid Electrical energy.

[0075] 2) Constraints on power generation and power purchase and sale These include constraints on the capacity of the main power grid interconnection lines, the upper limit of wind power output, the upper limit of photovoltaic power output, and the upper limit of gas turbine output.

[0076] The constraints on power generation and power purchase and sale specifically include:

[0077]

[0078]

[0079] In the formula, , , These are the upper limits for electricity purchase, photovoltaic power output, and wind power output, respectively.

[0080] 3) Energy storage system operation constraints This includes upper limits for energy storage charging power, upper limits for discharging power, mutual exclusion constraints for charging and discharging, constraints for updating the state of charge, upper and lower limits for the state of charge, and constraints for consistency of the state of charge at the beginning and end of the scheduling cycle.

[0081] The specific operational constraints of energy storage systems include:

[0082]

[0083]

[0084]

[0085] In the formula: This represents the maximum discharge power of the energy storage system. This is the maximum charging power of the energy storage system. and All are Boolean variables, serving as indicators of charging and discharging of the energy storage system; For the park middle The state of charge of the energy storage system at all times; , These refer to the charging and discharging efficiency of the energy storage system; , These are the upper and lower limits of the state of charge of the energy storage system, respectively. , The initial and final states of charge of the energy storage system; For the park Rated capacity of medium-sized energy storage system; To optimize the duration of a single time period during scheduling.

[0086] 4) Adjustable load operation constraints taking into account random response time delays The adjustable load operation constraints that take into account response time delay randomness include: load shift operation constraints that take into account response time delay randomness corresponding to time-shifted loads, power transfer constraints that take into account response time delay randomness corresponding to transferable loads, and power reduction constraints that take into account response time delay randomness corresponding to loads that can be reduced. The load shifting operation constraints that take into account the randomness of response time delay include allowable shifting time window, continuous operation constraint after startup, only one startup constraint, and actual operating power constraint under different response time delays. The power transfer constraints that take into account the randomness of response time delay include total power consumption conservation constraints, upper and lower limits of operating power constraints, minimum continuous operating time constraints, and start-up time constraints under different response time delays. The power reduction constraints that take into account the randomness of response delay include the allowable reduction time window, the maximum reduction ratio constraint, the minimum continuous reduction duration constraint, and the actual power reduction constraint under different response delays.

[0087] (1) Time-shifted load As a typical flexible adjustment resource, time-shifted loads are characterized by their ability to flexibly adjust the overall electricity consumption period to different times within a preset shiftable time window, thereby realizing the transfer of load in the time dimension.

[0088] The load shifting operation constraints corresponding to the time-shifted load, which take into account the randomness of response time delay, specifically include: This embodiment defines the time window during which time-shifted loads are allowed to perform power shifting within the scheduling cycle. Once the time-shifted load is started, the minimum continuous operating time required is Due to command response delays, the actual load startup time lags behind the command issuance time. Relevant operational constraints are explained below, including: Actual operating power constraints under different response delays:

[0089]

[0090] Constraints for continuous operation after startup:

[0091] Constraints are activated only once:

[0092] The load start-up identifier variable and the time-delay shift start-up state satisfy the following mapping relationship:

[0093] In the formula: For the park Middle, No In a typical response time delay scenario, time-shifted load... Adjusting power at any time; For the park middle Original power before time-shifted load optimization; For the park middle Total regulating power after time-shifted load optimization; K This represents the total number of typical response time delays; For the first The probability value corresponding to a typical response time delay; , These are Boolean variables, representing t, , and respectively. m The time-shifted load is in a power shifting state at any given moment. A value of 1 indicates that it is in a power shifting state, and a value of 0 indicates that it is not in a power shifting state. This is the minimum continuous operating time required once the time-shifted load is started; For the park i Time-shifted load in the first Response delay in a typical response delay scenario; It is a Boolean variable, representing the first... k In a typical response time delay scenario, the time-shifted load at time... The actual response trigger flag is set to 1 if the response is triggered and 0 if the response is not triggered. This is a Boolean variable representing the time-shifted load at... t The control command issuance flag is set to 1 to indicate issuance and 0 to indicate non-issuance. , These are the time windows during which power shifting is permitted for time-shifted loads. The lower limit time and the upper limit time.

[0094] (2) Transferable load The power transfer constraints corresponding to the transferable load, taking into account the randomness of the response time delay, specifically include: Transferable loads offer high flexibility, allowing for free adjustment of their operating time within a specified time interval based on system demand throughout the entire scheduling cycle. To characterize their response behavior and prevent equipment wear and tear due to frequent start-ups and shutdowns, the relevant operating constraints are as follows: Total electricity consumption conservation constraint:

[0095]

[0096] Operating power upper and lower limit constraints:

[0097] Minimum continuous runtime constraint:

[0098] Startup time constraints under different response delays:

[0099] Instruction and time-delay start-trigger association constraints:

[0100] In the formula: For the park middle Total regulating power optimized for load transfer at any time; For the park middle The original power before load transfer optimization; , These are the time windows during which power transfer is permitted for transferable loads. The lower limit time and the upper limit time; For the first The probability value corresponding to a typical response time delay; For the park Middle, No In a typical response time delay scenario, the transferable load is Adjusting power at any time; , The transferable loads are respectively The upper and lower limits of power at any given time; , These are Boolean variables, representing t, , and respectively. m This indicates whether the transferable load is in operation at any given time; a value of 1 indicates that it is in operation, and a value of 0 indicates that it is not in operation. This is a Boolean variable, representing the transferable load. t The control command issuance flag is set to 1 to indicate issuance and 0 to indicate non-issuance. The shortest duration of the transferable load; For the park i Transferable load in the first Response delay in a typical response delay scenario; It is a Boolean variable, representing the first...k In a typical response time delay scenario, the transferable load at time... The actual response trigger flag is set to 1 if the response is triggered and 0 if the response is not triggered.

[0101] (3) Load can be reduced The power reduction constraints that take into account the randomness of response time delays corresponding to the load reduction capacity specifically include: Reduceable loads mainly refer to users proactively reducing their power consumption to respond to system dispatch, based on their own electricity usage plans and while ensuring basic electricity needs are met. To reasonably describe this type of load reduction behavior, a time window is set within the dispatch cycle during which reduceable loads are allowed to implement power reduction. Its operational constraints are set as follows, including: Actual operating power constraints that can reduce load:

[0102] Actual power reduction constraints under different response delays:

[0103] Minimum consecutive reduction duration constraint:

[0104] Maximum reduction ratio constraint:

[0105] This can reduce the start-up timing constraints on the actual load participation in the response:

[0106] Command and time delay reduction trigger associated constraints:

[0107] In the formula: For the park middle Load power after time reduction; For the park Middle, No In a typical response time delay scenario Reduce the load power at any time; For the park middle Load power before the moment of reduction; For the first The probability value corresponding to a typical response time delay; For the park middle The load reduction rate at any given time; , These are Boolean variables, representing respectively t , m This indicates whether the load reduction is in a state of operation at any given time; a value of 1 indicates that the load is in a state of operation ... load reduction; This represents the maximum load reduction rate. For the park i The load can be reduced in the first Response delay in a typical response delay scenario; It is a Boolean variable, representing the first... k In a typical response time delay scenario, the load can be reduced at time [time]. The actual response trigger flag is set to 1 if the response is triggered and 0 if the response is not triggered. This is a Boolean variable, representing the load that can be reduced. t The control command issuance flag is set to 1 if issued and 0 if not issued. , These are the time windows during which power reduction can be implemented, allowing for load reduction. The lower limit time and the upper limit time; This is the lower limit for sustained load reduction.

[0108] Step 4: Solve the scheduling optimization model to obtain a multi-park collaborative scheduling operation scheme.

[0109] The multi-park collaborative scheduling and operation scheme includes the electricity purchase and sale plan, energy storage charging and discharging plan, gas turbine output plan, inter-park power exchange plan, and various adjustable load adjustment plans for each park during each scheduling period.

[0110] The scheduling optimization model is a mixed integer programming model. Those skilled in the art can use mature solution algorithms such as branch and bound algorithm and cutting plane algorithm, or various solvers to solve it. This invention does not limit the specific solution method of the model. Any existing technical solution that can solve the optimization model can be applied to this invention.

[0111] Furthermore, based on the optimization results and actual implementation results, the operation scheduling, cost settlement, and carbon emission accounting of the park cluster will be completed.

[0112] Example 2: To verify the superiority of the proposed scheduling strategy, this invention constructed three integrated energy system examples at the park level. Each park is equipped with wind turbine generators, photovoltaic power generation units, gas turbines, and energy storage devices. The load and wind / solar output curves of each park within 24 hours are shown below. Figure 2As shown in Table 1, the adjustable load parameter settings are as follows. The rated charge / discharge power of the energy storage system is 240 kW, the energy storage capacity is 480 kWh, the initial SOC is set to 0.6, and the charge / discharge efficiency is 0.95, while considering the lifespan degradation cost of 0.5 yuan / kWh per unit of electricity. The rated output power of the gas turbine is set to 500 kW, and the consumption parameters are as follows. Yuan / (kWh) 2 , Yuan / (kWh), The scheduling cycle is 24 hours, and the time resolution is 15 minutes. The response delay when adjustable loads participate in demand response is set to 1 to 4 steps, with corresponding probabilities of 0.16, 0.34, 0.36, and 0.14, respectively. The value is 0.728; The data used is the carbon emission factor data for a city in Jiangsu Province throughout the day; and the established optimization scheduling model is calculated and solved.

[0113] Table 1 Adjustable load parameters

[0114] Simulation Result Analysis To verify the effectiveness and feasibility of the strategy proposed in this paper, two comparative schemes were set up for analysis.

[0115] Option 1: Adopt the strategy proposed in this paper.

[0116] Option 2: When formulating the scheduling strategy, the spatiotemporal asynchronous characteristics of the adjustable load response are not considered, and it is also used to simulate the actual operation of the microgrid.

[0117] Figure 3 The results of energy supply and demand optimization for each park under Scheme 1 are presented. From an overall operational perspective, each park achieved dynamic power balance within the park at different times through the coordinated regulation of renewable energy output, gas turbine power generation, energy storage system charging and discharging, and power purchase and sale with the external power grid. During the daytime, as photovoltaic output gradually increases, the proportion of renewable energy generation in the energy supply structure significantly increases, effectively meeting the park's load demand at certain times. Simultaneously, the energy storage system charges when renewable energy output is high to absorb surplus energy and reduce power curtailment. During the evening peak period, as renewable energy output decreases and load levels increase, the energy storage system switches to discharging operation, working in conjunction with gas turbine output and power purchase from the external power grid to share the load demand, thereby alleviating the supply-demand imbalance. In contrast, during periods of low load or high renewable energy output, some parks can also sell electricity to the external power grid, improving energy utilization efficiency and system economy. Simultaneously, adjustable loads participate in regulation through demand response, smoothing out power fluctuations caused by renewable energy output, making system operation more stable, and improving the operational stability and economy of the integrated energy system.

[0118] Figure 4 The figure illustrates the power interaction between the various parks under Scheme 1. As shown, the power injected from Park 1 into Park 2 is positive for most of the time and has a relatively large amplitude, indicating that Park 1 continuously supplies power to Park 2. The power injected from Park 1 into Park 3 exhibits a bidirectional fluctuation characteristic, with its value varying around zero. However, the power interaction between Park 2 and Park 3 is close to zero, indicating that there is little direct power exchange between them. This result suggests that Park 1 has sufficient renewable energy output and the ability to achieve regional power output.

[0119] Figure 5 The actual carbon emission changes of the two scheduling schemes are presented. As shown in the figure, compared to Scheme 2, Scheme 1 has lower carbon emissions for most periods, especially during the daytime when renewable energy output is relatively abundant. Scheme 1 improves the proportion of renewable energy consumption and reduces the output demand of fossil fuel units such as gas turbines by optimizing the scheduling strategy, thereby effectively reducing system carbon emissions. Furthermore, Scheme 1 achieves near-zero emission operation during some periods, indicating that it can fully utilize renewable energy and reduce the participation of high-carbon power sources through energy storage and adjustable load coordination. Overall, Scheme 1 achieves better low-carbon operation while ensuring the safe and stable operation of the system, demonstrating a clear advantage in carbon emission control.

[0120] The energy supply and demand optimization results for each park under Plan 2 are as follows: Figure 6 As shown in the figure, compared with Scheme 1, although each park in Scheme 2 still maintains the system power balance through renewable energy output, gas turbine power generation, energy storage charging and discharging, and power purchase and sale with the external grid, the overall energy coordination level is relatively low. During periods of high partial load or large fluctuations in renewable energy output, the system's dependence on gas turbines and power purchase from the external grid increases significantly, indicating a decrease in renewable energy absorption capacity. Simultaneously, the power injection amplitude in each park is large, and the energy complementarity and coordinated regulation effects between parks are not significant. Because the dispatch strategy does not consider the spatiotemporal asynchronous characteristics of adjustable loads, the effect of smoothing power fluctuations is also relatively limited. In contrast, Scheme 1, by considering the spatiotemporal asynchronous characteristics of adjustable load response, makes the load regulation process more consistent with actual operating characteristics, effectively improving renewable energy absorption capacity and system operational stability.

[0121] Figure 7 This demonstrates the power interaction between the various parks under Scheme 2. Figure 4The comparison shows that the power fluctuation trend between the various parks is almost identical to that of Scheme 1. However, compared with Scheme 1, the power exchange fluctuation between the parks is more obvious in Scheme 2, indicating that ignoring the spatiotemporal asynchronous characteristics of adjustable loads will weaken the system's regulation capability and make it difficult to fully leverage the advantages of multi-park collaborative scheduling.

[0122] Economic analysis: Table 2 compares the operating costs of each park under the two schemes. As shown in the table, Scheme 1 is significantly lower than Scheme 2 in terms of electricity purchase cost, energy storage operating cost, and gas turbine operating cost. Specifically, the electricity purchase cost of Scheme 1 is 5034.06 yuan, a decrease of 300.22 yuan (5.63%) compared to Scheme 2; the energy storage cost is 34.11 yuan, a decrease of 256.10 yuan (88.25%) compared to Scheme 2; and the gas turbine operating cost is 153.26 yuan, a decrease of 244.50 yuan (61.47%) compared to Scheme 2. Due to the consideration of the spatiotemporal asynchronous characteristics of adjustable load, Scheme 1's overall system operating cost is 9831.07 yuan, a reduction of 876.26 yuan compared to Scheme 2's 10707.33 yuan, representing an 8.18% improvement in economic efficiency.

[0123] Table 2 Cost Comparison of Various Options

[0124] Example 3: A microgrid cluster consisting of three park-level microgrids is constructed, with each park equipped with wind turbine generators, photovoltaic power generation units, gas turbines, and energy storage devices. The dispatch cycle is set to 24 hours, and the time resolution to 15 minutes. The energy storage system has a rated charge / discharge power of 240kW, an energy storage capacity of 480kWh, an initial state of charge of 0.6%, and a charge / discharge efficiency of 0.95, taking into account a lifespan degradation cost of 0.5 yuan / kWh. The rated output power of the gas turbine is set to 500kW.

[0125] Adjustable load parameters can be set as follows: 1. Time-shifted load 1: Continuous operation time is 2 hours, the allowable shift time window is 5:00~21:00, and the unit compensation coefficient is 0.1 yuan / kWh; 2. Time-shifted load 2: Continuous operation time is 3 hours, the allowable shift time window is 7:00~23:00, and the unit compensation coefficient is 0.1 yuan / kWh; 3. Transferable load: minimum duration 1 hour, allowable adjustment time window is 4:00~22:00, unit compensation coefficient is 0.15 yuan / kWh; 4. Load reduction: Duration is 2 hours, the allowable reduction time window is 5:00~22:00, and the unit compensation coefficient is 0.2 yuan / kWh.

[0126] The response time delay of the adjustable load is set to 1-4 scheduling steps, corresponding to probabilities of 0.16, 0.34, 0.36, and 0.14, respectively. Carbon trading parameters, main grid power purchase and sale prices, and forecast curves for park load, wind power, and photovoltaic output can be input according to the actual operating scenario. Then, the established model is solved using an optimization solver to obtain the power purchase and sale capacity, energy storage charging and discharging power, gas turbine output power, adjustable load adjustment plan, and inter-park power interaction plan for each park.

[0127] The results show that during the daytime when photovoltaic output is high, the energy storage system prioritizes absorbing surplus renewable energy; during the evening peak when renewable energy output decreases, the energy storage system works with the gas turbine and the main grid to meet the load demand; there is obvious energy mutual assistance behavior between parks, indicating that the established model can effectively explore the spatiotemporal complementary characteristics between parks.

[0128] Alternative implementation methods: 1. The number of parks is not limited to three; it can be expanded to two, four, or more parks. 2. Conventional units are not limited to gas turbines; they can also be replaced by dispatchable power sources such as diesel generators, micro gas turbines, and fuel cells. 3. The time-delay random variable of adjustable load can be modeled using other discrete distributions or separately by park area and load type; 4. The carbon trading module can be replaced by a joint mechanism of carbon quota constraints, tiered carbon trading, or green certificate trading; 5. The method can be implemented either through a centralized scheduling platform or through a hierarchical distributed collaborative scheduling platform; 6. The technical solution can be further extended to a corresponding scheduling system, control device, electronic device, and computer-readable storage medium.

[0129] This disclosure can be a system, method, and / or computer program product. A computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for causing a processor to implement various aspects of this disclosure.

[0130] Computer-readable storage media can be tangible devices capable of holding and storing instructions for use by an instruction execution device. Computer-readable storage media can be, for example—but not limited to—electrical storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of computer-readable storage media include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital multifunction disc (DVD), memory sticks, floppy disks, mechanical encoding devices, such as punch cards or recessed protrusions storing instructions thereon, and any suitable combination of the foregoing. The computer-readable storage media used herein are not to be construed as transient signals themselves, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (e.g., light pulses through fiber optic cables), or electrical signals transmitted through wires.

[0131] The computer-readable program instructions described herein can be downloaded from computer-readable storage media to various computing / processing devices, or downloaded via a network, such as the Internet, local area network, wide area network, and / or wireless network, to an external computer or external storage device. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards them to the computer-readable storage media in the respective computing / processing device.

[0132] Computer program instructions used to perform the operations of this disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, status setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The computer-readable program instructions may execute entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, such as programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), is personalized by utilizing the status information of the computer-readable program instructions to implement various aspects of this disclosure.

[0133] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the claims of the present invention.

Claims

1. A collaborative optimization method for a cluster of microgrids in a park considering the asynchronous spatiotemporal response of loads, characterized in that, The method includes the following steps: Step 1: Collect operational data of the microgrids in each park; Step 2: Based on the load response characteristics, the park load is divided into conventional load and adjustable load, and a response time delay stochastic model of the adjustable load is constructed; the response time delay stochastic model discretizes the random response time delay of the adjustable load into multiple typical response time delays; Step 3: Based on the response time-delay stochastic model of the adjustable load, and combined with the operation data of each park microgrid, construct a low-carbon economic dispatch optimization model for multi-park microgrid groups with the goal of minimizing the overall operating cost. The dispatch optimization model includes adjustable load operation constraints that take into account the response time-delay stochasticity. Step 4: Solve the scheduling optimization model to obtain a multi-park collaborative scheduling operation scheme.

2. The collaborative optimization method for microgrid groups in a park according to claim 1, characterized in that: In step 2, the adjustable load includes time-shifted load, transferable load, and load that can be reduced; the response time-delay stochastic model of the adjustable load is specifically as follows: in, The random response time delay for adjustable loads; The time delay discretization of the random response is the first... k A typical value represents the park Inner The first type of adjustable load corresponding to A typical response time delay; Indicates the park Inner The adjustable load class has a time delay of 100° in random response. The probability, The corresponding probability value; K This represents the total number of typical response time delays.

3. The collaborative optimization method for microgrid groups in a park according to claim 2, characterized in that: The probability value Specifically, the following method is used to determine the load response time delay: collect historical control and operation data of the park's adjustable load, and statistically analyze load response time delay samples, wherein the load response time delay samples are the duration data between the time when the control command is issued and the actual time when the load responds; determine the load response time delay interval based on the sample distribution range; The total number of typical response time delays was selected based on the accuracy requirements of the power grid's intraday dispatch calculation. K The load response time delay interval is divided into K Each sub-interval is used as a characteristic time point to represent all time-delay conditions within its sub-interval. The frequency of the load response time delay sample falling into each sub-interval is statistically analyzed, and the corresponding proportion is obtained after normalization. This proportion is used as the probability value of the typical response time delay value within the corresponding sub-interval. .

4. The collaborative optimization method for microgrid groups in a park according to claim 2, characterized in that: In step 3, the adjustable load operation constraints that take into account the randomness of response time delay include: load shifting operation constraints that take into account the randomness of response time delay for time-shifted loads, power transfer constraints that take into account the randomness of response time delay for transferable loads, and power reduction constraints that take into account the randomness of response time delay for loads that can be reduced.

5. The collaborative optimization method for microgrid groups in a park according to claim 4, characterized in that: The load shifting operation constraints corresponding to the time-shifted load, which take into account the randomness of response time delay, specifically include: in, For the park Middle, No In a typical response time delay scenario, time-shifted load... Adjusting power at any time; For the park middle Original power before time-shifted load optimization; For the park middle Total regulating power after time-shifted load optimization; K This represents the total number of typical response time delays; For the first The probability value corresponding to a typical response time delay; , These are Boolean variables, representing t, , and respectively. m The time-shifted load is in the power shifting operation state at any given moment. A value of 1 indicates that it is in the power shifting state, and a value of 0 indicates that it is not in the power shifting state. This is the minimum continuous operating time required once the time-shifted load is started; For the park i Time-shifted load in the first Response delay in a typical response delay scenario; It is a Boolean variable, representing the first... k In a typical response time delay scenario, the time-shifted load at time... The actual response trigger flag is set to 1 if the response is triggered and 0 if the response is not triggered. This is a Boolean variable representing the time-shifted load at... t The control command issuance flag is set to 1 if issued and 0 if not issued. , These are the time windows during which power shifting is permitted for time-shifted loads. The lower limit time and the upper limit time.

6. The collaborative optimization method for microgrid groups in a park according to claim 4, characterized in that: The power transfer constraints corresponding to the transferable load, taking into account the randomness of the response time delay, specifically include: in, For the park middle Total regulating power optimized for load transfer at any time; For the park middle The original power before load transfer optimization; , These are the time windows during which power transfer is permitted for transferable loads. The lower limit time and the upper limit time; For the first The probability value corresponding to a typical response time delay; For the park Middle, No In a typical response time delay scenario, the transferable load is Adjusting power at any time; , The transferable loads are respectively The upper and lower limits of power at any given time; , These are Boolean variables, representing t, , and respectively. m This indicates whether the transferable load is in operation at any given time; a value of 1 indicates that it is in operation, and a value of 0 indicates that it is not in operation. This is a Boolean variable, representing the transferable load. t The control command issuance flag is set to 1 if issued and 0 if not issued. The shortest duration of the transferable load; For the park i Transferable load in the first Response delay in a typical response delay scenario; It is a Boolean variable, representing the first... k In a typical response time delay scenario, the transferable load at time... The actual response trigger flag is set to 1 if the response is triggered and 0 if the response is not triggered.

7. The collaborative optimization method for microgrid groups in a park according to claim 4, characterized in that: The power reduction constraints that take into account the randomness of response time delays corresponding to the load reduction capacity specifically include: in, For the park middle Load power after time reduction; For the park Middle, No In a typical response time delay scenario Reduce the load power at any time; For the park middle Load power before the moment of reduction; For the first The probability value corresponding to a typical response time delay; For the park middle The load reduction rate at any given time; , These are Boolean variables, representing respectively t , m This indicates whether the load reduction is in a state of operation at any given time; a value of 1 indicates that the load is in a state of operation ... load reduction; This represents the maximum load reduction rate. For the park i The load can be reduced in the first Response delay in a typical response delay scenario; It is a Boolean variable, representing the first... k In a typical response time delay scenario, the load can be reduced at time [time]. The actual response trigger flag is set to 1 if the response is triggered and 0 if the response is not triggered. This is a Boolean variable, representing the load that can be reduced. t The control command issuance flag is set to 1 if issued and 0 if not issued. , These are the time windows during which power reduction can be implemented, allowing for load reduction. The lower limit time and the upper limit time; This is the lower limit for the duration of load reduction.

8. A collaborative optimization system for a microgrid cluster in a park, based on the method of any one of claims 1-7, considering the asynchronous spatiotemporal response of loads, comprising a data acquisition module, a response time-delay stochastic model construction module for adjustable loads, a low-carbon economic dispatch optimization model construction module for multi-park microgrid clusters, and a dispatch optimization model solving module, characterized in that: The data acquisition module collects operational data from the microgrids in each park. The module for constructing a response time-delay stochastic model for adjustable loads divides the park load into conventional loads and adjustable loads according to the load response characteristics, and constructs a response time-delay stochastic model for adjustable loads; the response time-delay stochastic model discretizes the random response time delay of adjustable loads into multiple typical response time delays; The module for constructing a low-carbon economic dispatch optimization model for multi-park microgrid groups, based on the response time-delay stochastic model of the adjustable load and combined with the operation data of each park microgrid, constructs a low-carbon economic dispatch optimization model for multi-park microgrid groups with the goal of minimizing the overall operating cost. The dispatch optimization model includes adjustable load operation constraints that take into account the response time-delay stochasticity. The scheduling optimization model solving module solves the scheduling optimization model to obtain a multi-park collaborative scheduling operation scheme.

9. A terminal, comprising a processor and a storage medium; characterized in that: The storage medium is used to store instructions; The processor is configured to operate according to the instructions to perform the steps of the method according to any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the program implements the steps of the method according to any one of claims 1-7.

Images