Microgrid intelligent scheduling method based on multi-energy collaborative optimization

Abstract

The invention discloses a micro-grid intelligent scheduling method based on multi-energy collaborative optimization, and relates to the technical field of micro-grid scheduling, and the method comprises the steps: obtaining the subunit operation state information of an energy storage unit, and constructing a basic state parameter set for representing the current operation condition of the energy storage unit based on the subunit operation state information; the scheduling state of the energy storage unit in the current scheduling period is judged based on the basic state parameter set, a scheduling state identifier is generated, and the scheduling state identifier is a unique value in a single scheduling period and is used for limiting that the energy storage unit is only allowed to participate in one scheduling behavior type in the scheduling period. According to the method, the basic state parameter set is constructed and the state parameters representing the current operation conditions are extracted, so that the power capability, the energy boundary and the operation limitation of the energy storage unit in the current scheduling period can be accurately described, and a credible basis is provided for subsequent scheduling behavior identification and parameter selection.

Description

Technical Field

[0001] This invention relates to the field of microgrid dispatching technology, specifically to a microgrid intelligent dispatching method based on multi-energy collaborative optimization. Background Technology

[0002] With the increasing penetration rate of new energy sources, microgrids, as an important carrier of multi-energy integration, are playing an increasingly important role in improving energy utilization efficiency and enhancing the flexibility of the power system. Energy storage units in microgrids, due to their chargeable and dischargeable characteristics, have become a key supporting resource for regulating power balance and mitigating load fluctuations. In actual operation, intelligent scheduling and control of energy storage units has become an important research direction for achieving multi-energy synergistic optimization. Existing microgrid dispatching technologies mostly focus on optimal economic dispatching at the system level or local power balance control. They often drive dispatching decisions with global optimization models or pre-set fixed dispatching strategy execution processes. When dealing with energy storage dispatching behavior in complex and variable operating environments, these methods lack detailed characterization of the internal operating state of energy storage units, dispatching capacity boundaries, and cross-cycle operating consistency. This can easily lead to deviations between dispatching behavior and actual operating capacity, affecting system security and response effectiveness. On the other hand, with the diversification of scheduling behavior types, such as frequency regulation, energy transfer, and backup support, higher requirements are placed on the consistency control of the behavior of energy storage units within a single scheduling cycle. Currently, some scheduling mechanisms have introduced state judgment modules to assist in scheduling decisions, but they still lack a complete mechanism for selecting behavior types, constraining scheduling parameters, and performing dynamic consistency verification before command execution in the context of multi-source scheduling needs. This makes it difficult to fully release the regulation potential of energy storage resources and ensure execution stability. Therefore, this invention proposes a microgrid intelligent scheduling method based on multi-energy collaborative optimization. Summary of the Invention

[0003] The purpose of this invention is to provide a microgrid intelligent scheduling method based on multi-energy collaborative optimization to solve the problems mentioned in the background art.

[0004] This invention can be achieved through the following technical solution: a microgrid intelligent scheduling method based on multi-energy collaborative optimization, the method comprising: Step 1: Obtain the sub-unit operating status information of the energy storage unit, and construct a set of basic state parameters to characterize the current operating conditions of the energy storage unit based on the sub-unit operating status information; Step 2: Determine the scheduling status of the energy storage unit in the current scheduling cycle based on the set of basic state parameters, and generate a scheduling status identifier. The scheduling status identifier has a unique value in a single scheduling cycle and is used to limit the energy storage unit to only participating in one type of scheduling behavior in that scheduling cycle. Step 3: Based on the scheduling status identifier generated in Step 2, obtain multiple scheduling requirements for energy storage units, and filter them according to the scheduling status identifier, retaining the scheduling requirements that are consistent with the scheduling behavior type defined by the scheduling status identifier, thus forming a set of scheduling target parameters. Step 4: Under the constraints of the basic state parameter set, perform constraint processing on the scheduling target parameter set, determine the scheduling parameter values ​​that satisfy the current operating conditions, and generate scheduling instructions based on the scheduling parameter values; Step 5: While generating scheduling instructions, record the status identifier of the scheduling status identifier and the status identifier information of the basic status parameter set, and associate the status identifier information with the scheduling instructions; Step Six: Before executing the dispatch instruction, reacquire the sub-unit operating status information of the energy storage unit, update the basic status parameter set, and re-determine the dispatch status identifier; if the updated dispatch status identifier or basic status parameter set is inconsistent with the status identifier information associated in Step Five, prevent the dispatch instruction from entering the execution stage; if they are consistent, allow the dispatch instruction to enter the execution stage.

[0005] A further technical improvement of the present invention lies in: the method for constructing the basic state parameter set in step one, comprising: Within a preset acquisition period, the operating status information of each sub-unit of the energy storage unit is acquired, and the operating status information is time-aligned to form a set of sub-unit status data corresponding to the same acquisition time. The sub-unit operating status information includes at least the operating permit status, isolation identifier status, charge status, and power allowable range; The validity of each sub-unit status data set is determined. At least based on the operation license status and isolation identifier status, valid sub-unit status data that can be used to construct the basic status parameter set is identified, and sub-unit status data that do not meet the operation license conditions or are in an isolation state are excluded. Based on the effective sub-unit state data, the current operating conditions of the energy storage unit are aggregated and calculated. At least the corresponding available capacity boundary, available power boundary and continuous operation limit are collected, and a set of basic state parameters is generated according to the preset mapping relationship. State identification information is generated for the basic state parameter set, and the state identification information is associated with and stored in the basic state parameter set so that the basic state parameter set can be used as the input basis for subsequent scheduling state determination.

[0006] A further technical improvement of the present invention is that step two, generating the scheduling status identifier, includes: S1. Extract multiple decision dimension parameters from the basic state parameter set. The decision dimension parameters include the available power range, state of charge, remaining continuous operating time and maximum allowable power change rate in the current scheduling cycle. S2. Establish a set of judgment conditions corresponding to each scheduling behavior type based on the judgment dimension parameters; S3. Based on the judgment dimension parameters extracted in S1, predict the impact of the energy storage unit on the key state parameters in the subsequent preset number of scheduling cycles when it performs different scheduling behavior types in the current scheduling cycle. The prediction includes conservative linear extrapolation of the key state parameters based on the current state of charge, available power range and maximum allowable power change rate, to obtain the predicted state parameter set. S4. Compare the set of predicted state parameters with the state constraints corresponding to each scheduling behavior type. When the predicted state parameters corresponding to any scheduling behavior type exceed the limit of its state constraints, mark the scheduling behavior type as an invalid scheduling behavior type in the current scheduling cycle. S5. Among the remaining scheduling behavior types that were not marked as invalid in S4, according to the preset priority order, each scheduling behavior type is compared to see if it meets all the judgment conditions in the corresponding judgment condition set in S2. The scheduling behavior type that first meets the corresponding judgment condition is determined as the target scheduling behavior type, and the target scheduling behavior type is mapped to generate a scheduling status identifier so that the scheduling status identifier has a unique value in a single scheduling cycle.

[0007] A further technical improvement of the present invention is that the set of judgment conditions includes: a dynamic response capability judgment condition for identifying whether the ability to perform power regulation behavior is available; an energy margin judgment condition for identifying whether the ability to perform energy allocation behavior is available; and a safety judgment condition for identifying whether to enter a non-scheduling state when neither of the above two types of behavior meets the corresponding condition.

[0008] A further technical improvement of the present invention is that the method for forming the scheduling target parameter set in step three includes: Obtain all scheduling demand data for energy storage units within the current scheduling cycle; The scheduling demand data includes the scheduling behavior type identifier, demand response time, power or energy value, and duration. Based on the scheduling status identifier generated in step two, the scheduling requirements whose scheduling behavior type identifier is consistent with the scheduling status identifier are selected from the scheduling requirement data. The time validity of the filtered scheduling requests within the scheduling period is judged, and invalid requests whose response time is not within the current scheduling period are removed. The scheduling requirements that satisfy behavior type consistency and time validity are integrated into a set of scheduling target parameters, which serve as input parameters for subsequent generation of scheduling instructions.

[0009] A further technical improvement of the present invention is that step four, which involves constraint processing, includes: Based on the available power range, remaining continuous operating time, maximum allowable power change rate, and charge state parameters of the current scheduling cycle in the basic state parameter set, the feasible scheduling parameter range corresponding to each type of scheduling behavior is determined. The feasible scheduling parameter range represents the combination of scheduling parameters that can be accepted under the current state constraints. Map the power or energy value and duration parameter included in each scheduling requirement in the set of scheduling target parameters formed in step three to the corresponding feasible scheduling parameter range, and determine whether it falls within the range. For scheduling requirements that fall within the range of feasible scheduling parameters, calculate their remaining capacity margin relative to the boundary of feasible scheduling parameters under the current state conditions. Select the scheduling demand item with the maximum remaining capacity margin, extract its power or energy value and duration parameter, determine the final scheduling parameter value, and generate scheduling instructions based on the scheduling parameter value.

[0010] A further technical improvement of the present invention is that the calculation step of the residual capacity margin includes: Obtain the available power range, state of charge, remaining continuous operating time, and maximum allowable power change rate corresponding to the start time of the current scheduling cycle from the set of basic state parameters. Based on the duration parameter of each scheduling requirement in the scheduling target parameter set, and according to the maximum allowable power change rate and the state of charge change rate limit, taking the start time of the current scheduling cycle as the benchmark, the state parameters of the scheduling requirement in the covered multiple consecutive subsequent scheduling cycles are linearly numerically extrapolated to form the corresponding predicted state parameter sequence. For each periodic state parameter in the predicted state parameter sequence, compare it with the preset state safety lower limit parameter, and calculate the margin value of each periodic state parameter above the state safety lower limit parameter. The margin value is used to reflect the state safety margin of the scheduling requirement in the corresponding scheduling period. Based on the cycle number of each scheduling cycle in the predicted state parameter sequence, a set of time weighting coefficients that monotonically decreases as the cycle number increases is constructed. The margin value corresponding to each scheduling cycle is multiplied with its corresponding time weighting coefficient, and the product results of each scheduling cycle are accumulated to obtain a stability evaluation value that characterizes the degree of influence of the scheduling demand on the state safety boundary in multiple future scheduling cycles. The stability evaluation value is combined with the remaining capacity margin obtained from the boundary distances of power, energy, and duration dimensions within the current scheduling cycle to form the comprehensive remaining capacity margin value corresponding to the scheduling requirement.

[0011] A further technical improvement of the present invention is that step five, which associates the status identification information with the scheduling instruction, includes: During the process of generating scheduling instructions, the version identifiers corresponding to the scheduling status identifiers and basic status parameter sets generated in the current scheduling cycle are extracted simultaneously. Combine the scheduling status identifier, version identifier, and unique identifier of the scheduling instruction to construct a scheduling mapping record item, and write it into a structured scheduling mapping table; The scheduling mapping table includes a combined index key built based on the scheduling status identifier and the version identifier, and a timestamp field that records the generation time of the scheduling instruction. It is used to associate the scheduling instruction with the scheduling status and basic status parameters corresponding to the time of its generation.

[0012] A further technical improvement of the present invention is that step six, which involves performing a consistency check on the updated set of basic state parameters, includes: In the updated set of basic state parameters, extract multiple state determination parameters that are consistent with the determination dimension parameters defined in the scheduling state identifier generation rule. The determination dimension parameters include the available power range, state of charge, remaining continuous operating time, and maximum allowable power change rate corresponding to the current scheduling cycle. Calculate the change magnitude between each state determination parameter and its corresponding parameter value in the previously recorded basic state parameter set. Based on the constraint weight of each determination dimension parameter in the scheduling state identifier generation logic, the change magnitude of each dimension is weighted and superimposed to obtain the state change metric between the current basic state parameter set and the original basic state parameter set. The state change metric is compared with the preset state change threshold. When the state change metric exceeds the preset threshold, the scheduling state determination logic is re-executed, and an updated scheduling state identifier for the current scheduling cycle is generated. The updated scheduling status flag is compared with the historical scheduling status flag. If the two are inconsistent, it is determined that the scheduling premise of the current scheduling instruction is no longer valid, and the scheduling instruction is prevented from entering the execution phase.

[0013] Compared with the prior art, the present invention has the following beneficial effects: This invention constructs a set of basic state parameters and extracts state parameters that represent the current operating conditions, which can accurately characterize the power capacity, energy boundary and operating limitations of the energy storage unit in the current scheduling cycle, providing a reliable basis for subsequent scheduling behavior identification and parameter selection, and realizing dynamic scheduling input modeling that is strongly correlated with the operating state; Furthermore, this invention employs a unique scheduling status identification mechanism to ensure that only one type of scheduling behavior is allowed for an energy storage unit within a single scheduling cycle. Based on this identification, multiple scheduling requirements are filtered to ensure that the generated scheduling instructions match the current state conditions in terms of behavior type. At the same time, feasible parameter range constraints and remaining capacity margin evaluation indicators are introduced during the selection of scheduling parameters, which effectively improves the security and execution flexibility of the scheduling results. On the other hand, this invention establishes a structured mapping relationship between status identifiers and scheduling instructions after the scheduling instructions are generated, and introduces a two-factor consistency verification mechanism before execution to ensure the synchronization and adaptability of instruction execution with the current running state; when changes in status parameters cause the scheduling premise to fail, the system can automatically prevent instruction execution, thereby ensuring the consistency of scheduling behavior and the stability of system operation in a dynamic environment. Attached Figure Description

[0014] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0015] Figure 1 This is a schematic diagram of the method flow of the present invention. Detailed Implementation

[0016] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided.

[0017] Please see Figure 1 As shown, this invention provides a microgrid intelligent scheduling method based on multi-energy collaborative optimization, comprising: Step 1: Obtain the sub-unit operating status information of the energy storage unit, and construct a set of basic state parameters to characterize the current operating conditions of the energy storage unit based on the sub-unit operating status information. By collecting the operating status information of multiple sub-units within the energy storage unit during the current scheduling cycle, including but not limited to basic data such as operating permit status, isolation identifier status, state of charge, and power allowable range, the operating status data of effective sub-units are aggregated to form a set of basic state parameters, thereby accurately reflecting the operating capability boundary and state characteristics of the energy storage unit at the current moment, serving as the core input basis for subsequent stages of the scheduling logic.

[0018] Specifically, in this embodiment, to achieve dynamic constraints and accurate identification of the scheduling capability of the energy storage unit, the operating status information of each sub-unit in the energy storage unit is first periodically collected. The collection period can be set to 1-10 seconds according to the scheduling accuracy requirements. Within each scheduling period, the operating data of all sub-units are acquired through synchronous or quasi-synchronous methods.

[0019] Considering the time offset in the data acquisition of each sub-unit, the system needs to perform unified time benchmark alignment after completing the raw data retrieval. A time window buffering mechanism is used to align the data of the leading or lagging sub-units to ensure that a set of state data that can be uniformly parsed is formed on the same sampling node.

[0020] Subsequently, to ensure the validity of the input data and the security of the scheduling foundation, the validity of each sub-unit in the status data set needs to be judged. The judgment criteria are based on the operation permission status and the isolation status. The operation permission status can be provided by the protection device or energy management interface, indicating whether the current sub-unit is qualified to operate; the isolation status is used to determine whether the sub-unit has been physically disconnected from operation due to electrical faults, maintenance, or other reasons. The system only retains the sub-unit data with an operation permission status of "permitted" and an isolation status of "not isolated" as valid input.

[0021] For the effective set of sub-units, the system needs to perform aggregated modeling of their current capabilities under multiple dimensions such as electrical, thermal control, and energy. Specifically, firstly, the available capacities of all sub-units are summed to form the available capacity boundary at the energy storage unit level. Secondly, during the power aggregation process, maximum current constraints and bus capacity limitations are considered to form the upper and lower power limits of the energy storage unit in the charging and discharging directions. At the same time, historical information on thermal state and continuous operating time is introduced to calculate the continuous operating time limit. This parameter reflects the longest duration that the energy storage unit can sustain high-power operation under current heat dissipation capacity and historical load conditions.

[0022] The parameter set formed by the above aggregation will serve as the basis for determining the scheduling status in this cycle. To achieve its reusability and status traceability, the system generates structured identification information for this parameter set. The identification information is generated by combining the fields in the basic state parameter set according to a preset field order and encoding them based on a mapping function, thus generating state identification information that uniquely represents the current basic state parameter set. The identification information and the basic state parameter set are stored together in an indexable database for use in scheduling status identification and consistency verification.

[0023] Step 2: Determine the scheduling status of the energy storage unit within the current scheduling cycle based on the basic state parameter set, and generate a scheduling status identifier. The scheduling status identifier has a unique value within a single scheduling cycle, which is used to limit the energy storage unit to only participating in one type of scheduling behavior within that scheduling cycle. By extracting multiple judgment dimension parameters from the basic state parameter set and combining them with a preset set of judgment conditions, the scheduling behavior types that the energy storage unit can currently execute are identified, and a unique scheduling status identifier is generated based on the behavior type mapping. This ensures that the energy storage unit executes only one behavior mode within a scheduling cycle, thereby improving the consistency and response stability of scheduling behavior and avoiding role conflicts or behavior overlap issues.

[0024] Step two, generating the scheduling status identifier, includes: S1. Extract multiple decision dimension parameters from the basic state parameter set. The decision dimension parameters include the available power range, state of charge, remaining continuous operating time and maximum allowable power change rate in the current scheduling cycle. S2. Establish a set of judgment conditions corresponding to each scheduling behavior type based on the judgment dimension parameters; S3. Based on the judgment dimension parameters extracted in S1, predict the impact of the energy storage unit on the key state parameters in the subsequent preset number of scheduling cycles when it performs different scheduling behavior types in the current scheduling cycle. The prediction includes conservative linear extrapolation of the key state parameters based on the current state of charge, available power range and maximum allowable power change rate, to obtain the predicted state parameter set. S4. Compare the set of predicted state parameters with the state constraints corresponding to each scheduling behavior type. When the predicted state parameters corresponding to any scheduling behavior type exceed the limit of its state constraints, mark the scheduling behavior type as an invalid scheduling behavior type in the current scheduling cycle. S5. Among the remaining scheduling behavior types that were not marked as invalid in S4, according to the preset priority order, each scheduling behavior type is compared to see if it meets all the judgment conditions in the corresponding judgment condition set in S2. The scheduling behavior type that first meets the corresponding judgment condition is determined as the target scheduling behavior type, and the target scheduling behavior type is mapped to generate a scheduling status identifier so that the scheduling status identifier has a unique value in a single scheduling cycle.

[0025] Specifically, multiple dimensional parameters for judgment are extracted from the constructed set of basic state parameters. These judgment dimensional parameters include: the available power range corresponding to the current scheduling cycle, the current state of charge, the remaining continuous operating time, and the maximum allowable power change rate. All of these judgment dimensional parameters can be obtained by aggregating the effective sub-unit operating status data of the energy storage unit and mapping them through aggregation processing.

[0026] Based on the operational mechanism characteristics of various scheduling behavior types, a corresponding set of judgment conditions is established. This set of judgment conditions includes, but is not limited to: dynamic response capability judgment conditions for identifying whether the system has the ability to perform power regulation behavior; energy margin judgment conditions for identifying whether the system has the ability to perform energy allocation behavior; and safety judgment conditions for identifying whether the system needs to enter a non-scheduling state when neither of the above two types of behavior is met. For example, when the maximum allowable power change rate is not less than 50 kW / min and the upper boundary of the available power range is greater than 200 kW, the dynamic response capability judgment condition can be considered met; while when the state of charge is between 0.4 and 0.85 and the remaining continuous operating time exceeds 20 minutes, the energy margin judgment condition can be considered met.

[0027] To further evaluate the impact of different scheduling behaviors performed by energy storage units in the current scheduling cycle on future state security, based on extracted decision dimension parameters, a conservative prediction is made of the evolution trend of key state parameters in several subsequent preset scheduling cycles if the energy storage unit performs power regulation or energy allocation behaviors in the current scheduling cycle. During the prediction process, specific scheduling parameters are not directly used; instead, a set of typical scheduling parameters is set for each type of scheduling behavior for estimation. For example, to assess the feasibility of power regulation, 80% of the upper limit of the current available power range (e.g., 160kW) can be selected as the regulation power value, with a duration set to 3 minutes; for energy allocation, a typical constant charging power of -100kW can be selected. The prediction uses a linear extrapolation method, taking the current state of charge, available power range, and maximum allowable power change rate as inputs to generate a set of predicted state parameters covering several future scheduling cycles.

[0028] The predicted state parameter set is compared with the state constraints corresponding to each scheduling behavior type to determine whether the behavior is safe and feasible in future state evolution. State constraints may include: the state of charge must not be lower than 0.6 in all future prediction periods, and the predicted power value must not exceed the boundary determined by the current power upper limit and the allowable rate of change, etc. When the predicted state parameters of a certain behavior type exceed the limit of its constraints in any prediction period, the behavior type is marked as an invalid scheduling behavior type in the current scheduling period.

[0029] Among the scheduling behavior types not marked as invalid, each type is evaluated according to a preset priority order to determine whether it meets the corresponding set of judgment conditions, thus identifying the final target scheduling behavior type. The priority order is set through preset rules to ensure the uniqueness and interpretability of the decision. In a preferred embodiment, the decision rule is as follows: First, it is determined whether the power regulation behavior is not marked as invalid and meets the dynamic response capability judgment condition. If it does, it is selected as the target scheduling behavior type. If it does not, it is determined whether the energy allocation behavior meets its judgment condition and is not marked as invalid. If it does, it is determined as the target scheduling behavior type. If neither condition is met, the non-participation scheduling behavior is selected as the target scheduling behavior type. Through the above comparison process, it is ensured that the scheduling behavior types participated in by the energy storage unit in the current scheduling cycle have physical feasibility, control safety, and behavioral uniqueness. A scheduling status identifier is generated based on the target scheduling behavior type, and the scheduling status identifier has a unique value in the current scheduling cycle.

[0030] Step 3: Based on the scheduling status identifier generated in Step 2, obtain multiple scheduling requirements for energy storage units, and filter them according to the scheduling status identifier, retaining scheduling requirements that are consistent with the scheduling behavior type defined by the scheduling status identifier, forming a set of scheduling target parameters; by calling all available scheduling requirement data in the current cycle, and based on the scheduling status identifier generated in Step 2, perform consistency filtering on the behavior type in the scheduling requirement data, removing scheduling requirements with mismatched behavior types or not within the effective time window of the current cycle, and retaining only the set of scheduling target parameters that conform to the current status, thereby ensuring that the requirements involved in the scheduling calculation have instruction executability and behavior adaptability.

[0031] The methods for forming the scheduling target parameter set in step three include: Obtain all scheduling demand data for energy storage units within the current scheduling cycle; The scheduling demand data includes the scheduling behavior type identifier, demand response time, power or energy value, and duration. Based on the scheduling status identifier generated in step two, the scheduling requirements whose scheduling behavior type identifier is consistent with the scheduling status identifier are selected from the scheduling requirement data. The time validity of the filtered scheduling requests within the scheduling period is judged, and invalid requests whose response time is not within the current scheduling period are removed. The scheduling requirements that satisfy behavior type consistency and time validity are integrated into a set of scheduling target parameters, which serve as input parameters for subsequent generation of scheduling instructions.

[0032] Specifically, at the start of the current scheduling cycle, all scheduling demand data for the energy storage unit is acquired. This scheduling demand data can originate from the load response platform, energy management commands, or the local scheduling strategy generation module, and includes multiple fields: scheduling behavior type identifier, demand response time, power or energy value, and duration parameter. For example, a scheduling demand item might have a behavior type identifier of power regulation, a response time of 15:01, a power value of +120kW, and a duration of 5 minutes.

[0033] Based on the scheduling status identifiers generated in the previous scheduling step, scheduling requirements whose scheduling behavior type identifiers match the scheduling status identifiers are selected from the aforementioned scheduling requirement data. The scheduling status identifier limits the current energy storage unit to only one type of scheduling behavior, for example, only allowing energy allocation. Under this limitation, only requirements with the scheduling behavior type identifier of energy allocation are retained; other types of scheduling requirements are directly excluded. For example, in a cycle, if the scheduling status identifier is limited to energy allocation, and a requirement is of the power regulation behavior type, it will be filtered out even if the response time meets the requirements of the current scheduling cycle.

[0034] The scheduling requests that have passed the behavior type consistency screening are then evaluated for time validity. This evaluation compares the response time field recorded in the scheduling request with the time window of the current scheduling cycle. A scheduling request is considered valid only if its response time falls within the time range of the current scheduling cycle. For example, if the time interval of the current scheduling cycle is 15:00-15:05, only scheduling requests with response times within this time period are retained; requests with response times earlier than 15:00 or later than 15:05 are discarded to ensure the immediacy and consistency of the generated scheduling instructions.

[0035] The scheduling requirements that simultaneously satisfy both the consistency of scheduling behavior type and time validity are integrated to form the final set of scheduling target parameters. This set contains multiple candidate scheduling requirements, each with a clearly defined power or energy value, duration, and executability guarantee. The set of scheduling target parameters will serve as the input data basis for subsequent constraint processing steps, used to complete the selection of scheduling parameters and the generation of scheduling instructions within the feasible parameter space defined by the basic state parameter set.

[0036] This hierarchical screening and validity verification method ensures that the selection of scheduling parameters strictly follows the behavioral capability boundaries defined by the current scheduling state and system state, thereby improving the physical consistency and actual reachability of scheduling behavior and reducing the possibility of behavioral conflicts and invalid responses.

[0037] Step 4: Under the constraints of the basic state parameter set, the scheduling target parameter set is constrained to determine the scheduling parameter values ​​that meet the current operating conditions, and scheduling instructions are generated based on the scheduling parameter values. By utilizing the state constraint information such as power boundary, energy capacity, state of charge, and duration in the basic state parameter set, a feasible parameter range corresponding to the scheduling behavior is constructed. The range of each requirement parameter in the scheduling target parameter set is verified and the margin is evaluated. The most suitable combination of scheduling parameters for the current operating state is selected as the final scheduling parameter value, and a structured scheduling instruction is generated accordingly to drive the energy storage unit to perform scheduling operations.

[0038] Step four, the constraint processing steps, include: Based on the available power range, remaining continuous operating time, maximum allowable power change rate, and charge state parameters of the current scheduling cycle in the basic state parameter set, the feasible scheduling parameter range corresponding to each type of scheduling behavior is determined. The feasible scheduling parameter range represents the combination of scheduling parameters that can be accepted under the current state constraints. Map the power or energy value and duration parameter included in each scheduling requirement in the set of scheduling target parameters formed in step three to the corresponding feasible scheduling parameter range, and determine whether it falls within the range. For scheduling requirements that fall within the range of feasible scheduling parameters, calculate their remaining capacity margin relative to the boundary of feasible scheduling parameters under the current state conditions. Select the scheduling demand item with the maximum remaining capacity margin, extract its power or energy value and duration parameter, determine the final scheduling parameter value, and generate scheduling instructions based on the scheduling parameter value.

[0039] The steps for calculating residual capacity margin include: Obtain the available power range, state of charge, remaining continuous operating time, and maximum allowable power change rate corresponding to the start time of the current scheduling cycle from the set of basic state parameters. Based on the duration parameter of each scheduling requirement in the scheduling target parameter set, and according to the maximum allowable power change rate and the state of charge change rate limit, taking the start time of the current scheduling cycle as the benchmark, the state parameters of the scheduling requirement in the covered multiple consecutive subsequent scheduling cycles are linearly numerically extrapolated to form the corresponding predicted state parameter sequence. For each periodic state parameter in the predicted state parameter sequence, compare it with the preset state safety lower limit parameter, and calculate the margin value of each periodic state parameter above the state safety lower limit parameter. The margin value is used to reflect the state safety margin of the scheduling requirement in the corresponding scheduling period. Based on the cycle number of each scheduling cycle in the predicted state parameter sequence, a set of time weighting coefficients that monotonically decreases as the cycle number increases is constructed. The margin value corresponding to each scheduling cycle is multiplied with its corresponding time weighting coefficient, and the product results of each scheduling cycle are accumulated to obtain a stability evaluation value that characterizes the degree of influence of the scheduling demand on the state safety boundary in multiple future scheduling cycles. The stability evaluation value is combined with the remaining capacity margin obtained from the boundary distances of power, energy, and duration dimensions within the current scheduling cycle to form the comprehensive remaining capacity margin value corresponding to the scheduling requirement.

[0040] Specifically, in this embodiment, in order to select the scheduling requirement with the highest execution safety under the current basic state conditions from multiple feasible scheduling requirements, a comprehensive residual capacity margin value evaluation mechanism is further introduced to quantify the state margin performance of the current scheduling behavior in the current period and several future scheduling periods, thereby realizing multi-period safety optimization of the scheduling strategy.

[0041] Obtain the key operating parameters corresponding to the start time of the current scheduling cycle from the basic state parameter set. The key operating parameters include the available power range, state of charge, remaining continuous operating time, and maximum allowable power change rate, which serve as the starting reference for state evolution prediction after executing scheduling behavior.

[0042] For each candidate scheduling requirement in the set of scheduling target parameters, its duration parameter is extracted. Combined with constraints of the maximum allowable power change rate and the rate of change of state of charge, starting from the current scheduling cycle, linear numerical extrapolation of the state parameters is performed sequentially over multiple subsequent scheduling cycles covered by the requirement, forming a predicted state parameter sequence. The predicted state parameter sequence includes the state of charge parameters, available power parameters, and remaining continuous operating time parameters for each scheduling cycle.

[0043] For each scheduling cycle in the aforementioned predicted state parameter sequence, three key dimensions are extracted: state of charge parameter, available power parameter, and remaining continuous operating time parameter, and each is compared with its corresponding preset safety lower limit parameter. For each dimension, the positive difference above its safety lower limit represents the safety margin value for that scheduling cycle. Further, the safety margin values ​​of the three dimensions are weighted and summed according to preset importance weight coefficients to obtain the cycle safety margin value corresponding to that scheduling cycle. The importance weight coefficients can be set based on historical experience or scheduling strategies; for example, in a frequency stability priority scenario, the power dimension has a greater weight than other dimensions.

[0044] Based on the cycle number of each scheduling cycle in the predicted state parameter sequence, a set of time-weighted coefficients is constructed that monotonically decreases with increasing cycle number, indicating that the importance of different cycles to overall stability gradually weakens over time. The cycle safety margin value corresponding to each scheduling cycle is multiplied by its corresponding time-weighted coefficient, and the product results of all cycles are accumulated to obtain a stability evaluation value that reflects the degree of influence of the scheduling demand on the state safety boundary in multiple future scheduling cycles.

[0045] The stability evaluation value mentioned above is fused with the remaining capacity margin calculated based on the boundary distances of power, energy, and duration dimensions within the current scheduling cycle. Specifically, evaluation weight coefficients can be set for the current cycle and future cycles, and the comprehensive remaining capacity margin value corresponding to the scheduling requirement is obtained through a weighted summation method. The larger the comprehensive remaining capacity margin value, the more sufficient the state safety margin of the scheduling requirement is in the current and future cycles, and the lower the execution risk. Finally, the scheduling requirement with the largest comprehensive remaining capacity margin value is selected as the target execution object for the current cycle, and the power or energy value and duration parameter are extracted accordingly to complete the generation of the scheduling instruction.

[0046] Step 5: While generating scheduling instructions, record the status identifier of the scheduling status identifier and the status identifier information of the basic status parameter set, and associate the status identifier information with the scheduling instructions. At the same time as the scheduling instructions are generated, simultaneously extract the status identification information used to constrain the current scheduling decision, namely the status identifier of the scheduling status identifier and the status identifier information of the basic status parameter set, and store it in a structured association with the currently generated scheduling instructions to build a scheduling mapping relationship. This is used for consistency verification before subsequent instruction execution and traceability analysis of operation strategies, thereby enhancing the system's ability to perceive state changes and the controllability of scheduling behavior.

[0047] Step five involves associating status identification information with scheduling instructions, including: During the process of generating scheduling instructions, the version identifiers corresponding to the scheduling status identifiers and basic status parameter sets generated in the current scheduling cycle are extracted simultaneously. Combine the scheduling status identifier, version identifier, and unique identifier of the scheduling instruction to construct a scheduling mapping record item, and write it into a structured scheduling mapping table; The scheduling mapping table includes a combined index key built based on the scheduling status identifier and the version identifier, and a timestamp field that records the generation time of the scheduling instruction. It is used to associate the scheduling instruction with the scheduling status and basic status parameters corresponding to the time of its generation.

[0048] Specifically, while determining the final scheduling parameter values ​​and generating scheduling instructions, the scheduling status identifiers already determined within the current scheduling cycle are extracted in real time. These status identifiers are generated based on the basic status parameter set and decision dimension parameters, uniquely identifying the types of scheduling behaviors allowed to be executed within that scheduling cycle. Furthermore, version identifiers corresponding to the basic status parameter set are extracted simultaneously. These version identifiers represent the uniqueness of the basic status parameter set and are typically generated by combining the parameter fields constituting the set in a preset order and performing feature calculations, ensuring that it corresponds to only one unique snapshot of the running conditions at any given time.

[0049] The extracted scheduling status identifier, version identifier, and unique identifier of the scheduling instruction are structurally combined to construct a complete scheduling mapping record. The unique identifier can be obtained by normalizing and hashing the parameter fields (including power, energy, duration, etc.) in the scheduling instruction. For example, a structured encoding function can be used to encode "charging +150kW +10 minutes" as ID_202501051230, ensuring that the instruction is unique and reproducible.

[0050] The completed scheduling mapping records are written into a predefined, structured scheduling mapping table, which serves as a data storage structure for managing scheduling history. To improve retrieval efficiency, the scheduling mapping table uses a composite index key based on the scheduling status identifier and version identifier to support quick location of scheduling records according to their status context. Additionally, the scheduling mapping records include a timestamp field to record the precise moment the scheduling instruction was generated, facilitating the tracking of the scheduling decision's creation time in subsequent status backtracking or instruction comparison scenarios.

[0051] Through the aforementioned structured recording process, a data foundation is formed for tracing, comparing, and verifying scheduling behavior, achieving a strong binding association between scheduling instructions and their generation background conditions (including scheduling state and basic state). Before scheduling execution, the system can perform validity verification based on the scheduling mapping table to prevent scheduling decision mismatch due to operational state deviations, thereby improving the security and adaptability of the scheduling process.

[0052] Step Six: Before executing the scheduling command, the sub-unit operating status information of the energy storage unit is reacquired, the basic status parameter set is updated, and the scheduling status identifier is re-determined. If the updated scheduling status identifier or basic status parameter set is inconsistent with the status identifier information associated in Step Five, the scheduling command is prevented from entering the execution phase; otherwise, the scheduling command is allowed to enter the execution phase. Step Six serves as a safety verification mechanism before the scheduling command is executed. By refreshing the operating status of the energy storage unit in real time, updating the basic status parameter set, and regenerating the corresponding scheduling status identifier, the system compares the current status with the status identifier information when the scheduling command was generated. If it is found that the status has changed and the scheduling preconditions have failed, the command will be prevented from executing, thereby effectively avoiding the risks of improper execution, resource overruns, or behavior mismatch caused by sudden state changes, and ensuring the timeliness and rationality of command execution.

[0053] Step six involves performing a consistency check on the updated set of basic state parameters, which includes: In the updated set of basic state parameters, extract multiple state determination parameters that are consistent with the determination dimension parameters defined in the scheduling state identifier generation rule. The determination dimension parameters include the available power range, state of charge, remaining continuous operating time, and maximum allowable power change rate corresponding to the current scheduling cycle. Calculate the change magnitude between each state determination parameter and its corresponding parameter value in the previously recorded basic state parameter set. Based on the constraint weight of each determination dimension parameter in the scheduling state identifier generation logic, the change magnitude of each dimension is weighted and superimposed to obtain the state change metric between the current basic state parameter set and the original basic state parameter set. The state change metric is compared with the preset state change threshold. When the state change metric exceeds the preset threshold, the scheduling state determination logic is re-executed, and an updated scheduling state identifier for the current scheduling cycle is generated. The updated scheduling status flag is compared with the historical scheduling status flag. If the two are inconsistent, it is determined that the scheduling premise of the current scheduling instruction is no longer valid, and the scheduling instruction is prevented from entering the execution phase.

[0054] Specifically, the updated set of basic state parameters is first obtained, and then multiple state judgment parameters consistent with the judgment dimension parameters defined in the scheduling state identifier generation rules are extracted. These judgment dimension parameters include: the available power range corresponding to the current scheduling cycle, the state of charge, the remaining continuous operating time, and the maximum allowable power change rate. This step ensures that the state information involved in the verification is consistent with the scheduling state judgment, thus ensuring comparability.

[0055] For each of the above state determination parameters, the magnitude of change between its value and the corresponding parameter value in the previously recorded set of basic state parameters is calculated, using the absolute difference method. For example, the magnitude of change in power range can be represented by the difference between the average values ​​of the upper and lower boundaries, the state of charge and remaining operating time can be expressed as the difference between actual values, and the maximum permissible rate of power change can be expressed as the absolute value of the rate of change.

[0056] Based on this, according to the importance of the constraints set in the scheduling state determination logic, corresponding weights are assigned to each state determination parameter. Examples of weights include: available power range weight is 0.35, state of charge weight is 0.35, remaining continuous operating time weight is 0.20, and maximum allowable power change rate weight is 0.10. The change magnitude of each dimension is multiplied by its weight and then summed in a weighted manner to obtain the state change metric between the current set of basic state parameters and the original set of basic state parameters.

[0057] The aforementioned state change metric is compared with a preset state change threshold to determine whether to trigger a regeneration of the scheduling status identifier. The state change threshold can be configured according to different operating scenarios, typically using 10% to 15% of the full scale of the basic state parameter set as a reference setting, for example, 0.12 under typical operating conditions. When the state change metric exceeds this threshold, it is determined that the current state has changed beyond the safety tolerance, thereby triggering the scheduling status reassessment logic and generating an updated scheduling status identifier for the current scheduling cycle.

[0058] Finally, the updated scheduling status identifier is compared with the historical scheduling status identifier. If they match, it indicates that the current running state is consistent with the premise state when the scheduling instruction was generated, and the scheduling instruction is allowed to enter the execution phase. If they do not match, it is considered that the current running state no longer meets the original scheduling conditions, and the scheduling instruction is prevented from being executed. This ensures that the system's running behavior matches the actual state and avoids scheduling risks caused by state mismatch.

[0059] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters and thresholds in the formulas are set by those skilled in the art according to the actual situation.

[0060] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A micro-grid intelligent scheduling method based on multi-energy collaborative optimization, characterized in that, The method comprises the following steps: Step one, obtaining the subunit operating state information of the energy storage unit, and constructing a basic state parameter set for representing the current operating condition of the energy storage unit based on the subunit operating state information; Step two, determining the scheduling state of the energy storage unit in the current scheduling period based on the basic state parameter set, and generating a scheduling state identifier, wherein the scheduling state identifier has a unique value in a single scheduling period, and is used to limit the energy storage unit to only participate in one type of scheduling behavior in the scheduling period; Step three, based on the scheduling state identifier generated in step two, obtaining multiple types of scheduling requirements for the energy storage unit, and filtering according to the scheduling state identifier to retain the scheduling requirements consistent with the scheduling behavior type defined by the scheduling state identifier, and forming a scheduling target parameter set; Step four, under the limitation of the basic state parameter set, the scheduling target parameter set is processed, the scheduling parameter value meeting the current operating condition is determined, and the scheduling instruction is generated based on the scheduling parameter value; Step five, while generating the scheduling instruction, record the state identifier information of the scheduling state identifier and the basic state parameter set, and associate the state identifier information with the scheduling instruction; Step six, before executing the scheduling instruction, the subunit operating state information of the energy storage unit is reacquired, the basic state parameter set is updated, and the scheduling state identifier is re-determined; when the updated scheduling state identifier or the basic state parameter set is inconsistent with the state identifier information associated in step five, the scheduling instruction is prevented from entering the execution stage, and when they are consistent, the scheduling instruction is allowed to enter the execution stage. 2.The micro-grid intelligent scheduling method based on multi-energy collaborative optimization according to claim 1, characterized in that, The construction method of the basic state parameter set in step one comprises: In a preset collection period, the operating state information of each subunit of the energy storage unit is obtained, and the operating state information is time-aligned to form a subunit state data set corresponding to the same collection time; The subunit operating state information at least includes operating permission state, isolation identifier state, state of charge and power allowable range; The validity of each subunit state data set is identified, and at least the operating permission state and the isolation identifier state are used to identify the valid subunit state data that can be used to construct the basic state parameter set, and the subunit state data that does not meet the operating permission condition or is in the isolation state is excluded; Based on the valid subunit state data, the current operating condition of the energy storage unit is aggregated and calculated, at least the corresponding available capacity boundary, available power boundary and continuous operation limit are collected, and the basic state parameter set is generated according to the preset mapping relationship; The state identifier information of the basic state parameter set is generated, and the state identifier information is stored in association with the basic state parameter set, so that the basic state parameter set can be used as the input basis for subsequent scheduling state determination. 3.The micro-grid intelligent scheduling method based on multi-energy collaborative optimization of claim 1, wherein, The step of generating the scheduling state identifier in step two comprises: S1, extracting multiple judgment dimension parameters from the basic state parameter set, the judgment dimension parameters including the available power range, the state of charge, the remaining continuous operation time and the maximum allowed power change rate in the current scheduling period; S2, establishing a judgment condition set corresponding to each scheduling behavior type based on the judgment dimension parameters; S3, based on the determination dimension parameters extracted in S1, the influence of the energy storage unit on the key state parameters in the subsequent preset number of scheduling periods when performing different scheduling behavior types in the current scheduling period is predicted, the prediction including conservatively linearly extrapolating the key state parameters according to the current state of charge, the available power range and the maximum allowed power change rate to obtain a set of predicted state parameters; S4, compare the set of predicted state parameters with the state constraint conditions corresponding to each scheduling behavior type, and when the predicted state parameters corresponding to any scheduling behavior type exceed the limited range of the state constraint condition, mark the scheduling behavior type as an invalid scheduling behavior type in the current scheduling period; S5, among the scheduling behavior types remaining in S4 which are not marked as invalid, according to the preset priority order, compare whether each scheduling behavior type meets all the judgment conditions in the corresponding set of judgment conditions in S2 in turn, determine the scheduling behavior type that first meets the corresponding judgment condition as the target scheduling behavior type, and map the target scheduling behavior type to generate a scheduling state identifier, so that the scheduling state identifier has a unique value in a single scheduling period.

4. The microgrid intelligent scheduling method based on multi-energy collaborative optimization according to claim 3, characterized in that, The set of judgment conditions includes: a dynamic response capability judgment condition for identifying whether it has the ability to perform power regulation behavior, an energy margin judgment condition for identifying whether it has the ability to perform energy distribution behavior, and a safety judgment condition for identifying whether it needs to enter a non-participation scheduling state when neither of the above two behaviors meets the corresponding condition.

5. The method of claim 1, wherein, The method of forming the scheduling target parameter set in step three includes: Obtain all scheduling demand data for the energy storage unit in the current scheduling period; The scheduling demand data includes scheduling behavior type identifier, demand response time, power or energy value and duration; Based on the scheduling state identifier generated in step two, filter the scheduling demand whose scheduling behavior type identifier is consistent with the scheduling state identifier from the scheduling demand data; Perform time validity judgment of the filtered scheduling demand within the scheduling period, and eliminate invalid demand whose response time is not within the current scheduling period; Integrate the scheduling demand that meets the behavior type consistency and time validity to form the scheduling target parameter set as the input parameter for generating the scheduling instruction subsequently. 6.The microgrid intelligent scheduling method based on multi-energy collaborative optimization of claim 1, wherein, The step of step four for constraint processing includes: Determine the feasible scheduling parameter range corresponding to each type of scheduling behavior type according to the available power range, the remaining continuous operation time, the maximum allowed power change rate and the state of charge parameter in the current scheduling period in the basic state parameter set, which represents the acceptable scheduling parameter combination under the current state constraint; Map the power or energy value and duration parameters included in each scheduling demand in the scheduling target parameter set formed in step three to the corresponding feasible scheduling parameter range, and judge whether it falls within the range; For the scheduling demand items falling within the feasible scheduling parameter range, calculate the remaining capability margin relative to the feasible scheduling parameter boundary under the current state condition; The scheduling demand item with the largest remaining capability margin is selected, the power or energy value and the duration parameter thereof are extracted, the final scheduling parameter value is determined, and a scheduling instruction is generated based on the scheduling parameter value.

7. The microgrid intelligent scheduling method based on multi-energy collaborative optimization according to claim 6, characterized in that, The step of calculating the remaining capability margin comprises: obtaining the available power range, the state of charge, the remaining continuous operation time and the maximum allowable power change rate corresponding to the starting moment of the current scheduling period in the basic state parameter set; based on the duration parameter of each scheduling demand in the scheduling target parameter set, linear numerical deduction is performed on the state parameters of the scheduling demand in the covered continuous subsequent scheduling periods from the starting moment of the current scheduling period as the reference, under the maximum allowable power change rate and the state of charge change rate limit, to form a corresponding predicted state parameter sequence; for each period state parameter in the predicted state parameter sequence, the state safety lower limit parameter is compared respectively, the margin value of each period state parameter higher than the state safety lower limit parameter is calculated, and the margin value is used to reflect the state safety margin of the scheduling demand in the corresponding scheduling period; a set of time weighting coefficients monotonically decreasing with the increase of the cycle serial number is constructed according to the cycle serial number of each scheduling period in the predicted state parameter sequence; the margin value corresponding to each scheduling period is multiplied by the time weighting coefficient corresponding thereto, and the product results of each scheduling period are accumulated to obtain the stability evaluation value representing the influence degree of the scheduling demand on the state safety boundary in the future multiple scheduling periods. Step five comprises: 8.The micro-grid intelligent scheduling method based on multi-energy collaborative optimization of claim 1, wherein, in the process of generating the scheduling instruction, the scheduling state identifier generated in the current scheduling period and the version identifier corresponding to the basic state parameter set are extracted synchronously; the scheduling state identifier, the version identifier and the unique identifier information of the scheduling instruction are combined to construct a scheduling mapping record item, and are written into a structured scheduling mapping table; the scheduling mapping table comprises a joint index key constructed based on the scheduling state identifier and the version identifier and a time stamp field recording the generation time of the scheduling instruction, which is used to realize the association between the scheduling instruction and the scheduling state and the basic state parameter corresponding to the generation time. Step six comprises: 9.The micro-grid intelligent scheduling method based on multi-capability collaborative optimization of claim 1, wherein, in the updated basic state parameter set, multiple state judgment parameters consistent with the judgment dimension parameters defined in the scheduling state identifier generation rule are extracted, the judgment dimension parameters including the available power range, the state of charge, the remaining continuous operation time and the maximum allowable power change rate corresponding to the current scheduling period; ​ The change amplitude between each state determination parameter and the corresponding parameter value in the previously recorded basic state parameter set is calculated respectively, and each dimension change amplitude is weighted and superimposed according to the constraint weight of each determination dimension parameter in the scheduling state identification generation logic to obtain the state change metric value between the current basic state parameter set and the original basic state parameter set; The state change metric value is compared with the preset state change threshold value, and when the state change metric value exceeds the preset threshold value, the scheduling state determination logic is re-executed, and the updated scheduling state identification of the current scheduling period is generated; The updated scheduling state identification is compared with the historically recorded scheduling state identification, and when the two are inconsistent, it is determined that the scheduling premise of the current scheduling instruction is no longer valid, and the scheduling instruction is prevented from entering the execution phase.

Images