Multi-generator unit-oriented collaborative operation and maintenance decision method and system

Abstract

The application provides a multi-generator set cooperative operation and maintenance decision method and system, relates to the technical field of data processing, and comprises the following steps: cutting a basic operation data set according to the same time granularity, calculating the coupling stability degree of the operation state of a generator set in an operation cycle to obtain a coupling strength value, merging the generator sets with coupling strength values higher than a preset coupling value into the same cooperative group, calculating the operation evolution stability degree of each cooperative group in the operation cycle to obtain an evolution stability value, giving each cooperative group a scheduling weight, mapping the scheduling weight into a unit start-stop priority and a load allocation ratio to obtain a cooperative scheduling scheme to generate an operation and maintenance decision instruction, and cooperatively operating and maintaining the multi-generator set according to the operation and maintenance decision instruction to obtain a cooperative operation and maintenance result. The application realizes the cooperative operation and maintenance of the multi-generator set through the real states of the multi-generator sets.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a collaborative operation and maintenance decision-making method and system for multiple generator sets. Background Technology

[0002] In the existing operation and maintenance management of multiple generator sets, independent monitoring and hierarchical scheduling are usually adopted to achieve operational coordination. Each generator set acquires operating parameters such as voltage, current, power and temperature, and uploads them to a centralized control platform. The dispatcher or fixed rule algorithm performs start-stop control and load allocation. When the system is running, it executes start-stop commands according to preset thresholds or schedules. When a unit is detected to be overloaded or underpowered, the system maintains power supply balance by switching to a standby unit or adjusting the output power. This rule-based scheduling method can achieve basic power supply stability and unit management in most fixed load scenarios.

[0003] However, in scenarios with multi-unit coordination, when external load demand fluctuates significantly or unit status changes frequently, the system may not be able to make adaptive decisions based on the health status, operating efficiency, and predicted output of each unit. For example, in a wind farm, when a unit's output power decreases due to blade contamination, the system may not be able to optimize the scheduling strategy by combining historical data with real-time status predictions if it only triggers the standby unit to start up based on a fixed threshold. This could lead to excessive start-up and shutdown of some units, uneven energy distribution, or overall power supply fluctuations, affecting the stability of system operation. Summary of the Invention

[0004] The purpose of this invention is to provide a collaborative operation and maintenance decision-making method and system for multiple generator sets, aiming to solve the problems mentioned in the background art.

[0005] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows:

[0006] Firstly, a collaborative operation and maintenance decision-making method for multiple generator units, the method comprising:

[0007] Acquire power change data, temperature change data, and load response data of the generator set within the same operating cycle to obtain a basic operating dataset;

[0008] The basic running dataset is divided into segments with the same time granularity, and the data within each time granularity is aligned and combined to obtain the time window running sequence;

[0009] Based on the time window operation sequence, the coupling stability of the generator set's operating state within the operating cycle is calculated to obtain the coupling strength value;

[0010] Generator sets with coupling strength values ​​higher than the preset coupling value are grouped into the same coordination group, and within the coordination group, they are arranged into a coordination group sequence according to the change range of load response data.

[0011] Based on the sequence of coordinating groups, the degree of operational evolution stability of each coordinating group within the operating cycle is calculated to obtain the evolution stability value;

[0012] Based on the evolutionary stability value, each coordination group is assigned a scheduling weight, and the scheduling weight is mapped to the unit start-up and shutdown priority and load allocation ratio to obtain the coordination scheduling scheme.

[0013] The operation and maintenance decision instructions are generated according to the collaborative scheduling scheme, and the collaborative operation and maintenance of multiple generator sets are carried out according to the operation and maintenance decision instructions to obtain the collaborative operation and maintenance results.

[0014] Furthermore, the basic runtime dataset is divided into segments with the same time granularity, and the data within each time granularity are aligned and combined to obtain the time window runtime sequence, including:

[0015] Based on the basic operation dataset, extract the time stamps for each power change data, temperature change data, and load response data, and organize the time stamps in the order of the operation cycle to obtain a time index set;

[0016] The time index set is divided into multiple consecutive time periods according to the same time granularity, and a segment number identifier is generated for each consecutive time period to obtain a time segmentation table.

[0017] Based on the time segmentation table, each data item in the basic operation dataset is assigned to the corresponding segment number according to its time identifier, and each data item is aggregated according to the unit number to obtain segmented aggregated data;

[0018] Based on the segmented aggregated data, each data item under the same segment number is matched according to the unit number, and the matching results are encapsulated into time window units to obtain time window combined data;

[0019] Arrange the time window combination data in ascending order according to the segment number identifier, and add a sequence position mark to each time window unit to obtain the time window running sequence.

[0020] Furthermore, based on the time window operating sequence, the coupling stability of the generator set's operating state within the operating cycle is calculated to obtain the coupling strength value, including:

[0021] Based on the time window operation sequence, calculate whether the power change direction and temperature change direction are consistent within each time granularity to obtain the direction consistency suppression term; calculate the interaction value between the power change amplitude and temperature change amplitude within each time granularity to obtain the amplitude interaction contribution term; calculate the degree of disturbance of the load response amplitude within each time granularity to obtain the load disturbance reduction term.

[0022] Based on the direction consistency suppression term, the continuity of the direction consistency state within two adjacent time granularities is calculated to obtain the adjacent continuous enhancement term; based on the direction consistency suppression term, amplitude interaction contribution term, and load disturbance reduction term, the coupling contribution value within each time granularity is calculated to obtain the coupling contribution merging term.

[0023] Based on the coupling contribution merging term and adjacent consecutive enhancement terms, the enhanced coupling contribution value within each time granularity is calculated to obtain the coupling strength value.

[0024] Furthermore, generator sets with coupling strength values ​​higher than a preset coupling value are grouped into the same coordination group, and within the coordination group, they are arranged into a coordination group sequence according to the variation amplitude of load response data, including:

[0025] Generator sets with coupling strength values ​​higher than a preset coupling value are marked as candidate generator sets, and an index is established between the generator set number and the coupling strength value of the candidate generator sets to obtain a set of candidate generator sets;

[0026] A merging rule table is generated based on the index association of the candidate unit set, and the candidate units are assigned to the cooperative group identifier according to the merging rule table. When the same candidate unit satisfies multiple cooperative group identifiers at the same time, the candidate unit is merged into the cooperative group identifier with the highest coupling strength value to obtain the cooperative group set.

[0027] Based on the set of collaborative groups, the load response data of each generator set in the collaborative group during the operating cycle is extracted, and the peak-valley difference of the load response data and the cumulative value of the granularity change of adjacent time are combined into the load amplitude characteristic value to obtain the load amplitude value within the group.

[0028] Based on the load amplitude value within the group, the generator sets within the coordinated group are sorted according to the magnitude of the load amplitude characteristic value to obtain the coordinated group sequence.

[0029] Furthermore, based on the sequence of coordinating groups, the operational evolution stability of each coordinating group within the operating cycle is calculated to obtain the evolution stability value, including:

[0030] Based on the cooperative group sequence, the mean values ​​of power change data and temperature change data of each generator set in the cooperative group within each time granularity are calculated and used as power reference trajectory and temperature reference trajectory to obtain the reference trajectory term within the group; the deviation of power change data of each generator set in the cooperative group from the power reference trajectory within each time granularity is calculated to obtain the dispersion term within the group.

[0031] Based on the intra-group reference trajectory term, the continuity of the evolution trajectory of the power reference trajectory and temperature reference trajectory between adjacent time granularities is calculated to obtain the intra-group smoothness term; the aggregation amplitude of the load response data of each unit in the cooperative group within each time granularity is calculated to obtain the disturbance gating term.

[0032] Based on the intra-group dispersion term, intra-group smoothness term, and disturbance gating term, the stable contribution value of each time granularity is calculated to obtain the single-time-window stable contribution term; based on the single-time-window stable contribution term and the disturbance gating term, the cumulative stable contribution value of all time granularities within the running cycle is calculated to obtain the evolutionary stable value.

[0033] Furthermore, based on the evolutionary stability value, scheduling weights are assigned to each coordination group, and these scheduling weights are mapped to unit start-up / shutdown priorities and load allocation ratios to obtain a coordinated scheduling scheme, including:

[0034] The evolutionary stability value of each coordination group is mapped to a scheduling weight, and the scheduling weight is continuously corrected according to the preset weight jump restriction rule to obtain the coordination group scheduling weight.

[0035] The scheduling weight of the coordination group is converted into the start-stop priority of the coordination group, and the group order in the coordination group sequence is inherited as the start-stop order of the group. At the same time, the start-stop interval constraint table is generated according to the coupling strength value of the units in the coordination group, and the start-stop interval constraint table is written into the start-stop priority table to obtain the unit start-stop priority.

[0036] The load demand data within the operating cycle is divided into load segments according to time granularity, and the load segments are proportionally allocated according to the scheduling weight of the coordination group. A proportional adjustment threshold table is generated based on the evolution and stability value of each coordination group to correct the proportional allocation results within the same time granularity, thus obtaining the load allocation ratio.

[0037] Furthermore, operation and maintenance decision instructions are generated based on the collaborative scheduling scheme, and collaborative operation and maintenance is performed on multiple generator units according to these instructions to obtain collaborative operation and maintenance results, including:

[0038] Based on the collaborative scheduling scheme, the unit start-up and shutdown priority, start-up and shutdown interval constraint table and load allocation ratio are extracted, and a parameter mapping relationship is established between them and the unit identifier according to the time granularity identifier to obtain the instruction parameter set.

[0039] The instruction parameter set is split into start / stop instruction fragments and load instruction fragments, and an execution order identifier is written for each instruction fragment to obtain the instruction fragment set;

[0040] Based on the instruction fragment set, conflict screening is performed on start / stop instruction fragments and load instruction fragments under the same time granularity identifier. The conflicting instruction fragments are then adjusted according to the start / stop interval constraint table and the total constraint of load allocation ratio to obtain the operation and maintenance decision instruction set.

[0041] Based on the operation and maintenance decision instruction set, the unit start-up and shutdown control instructions and load distribution control instructions are output sequentially, and the execution status of each time granularity is written into the operation and maintenance record table to obtain the collaborative operation and maintenance results.

[0042] Secondly, a collaborative operation and maintenance decision-making system for multiple generator sets, the system comprising:

[0043] The data module is used to acquire power change data, temperature change data, and load response data of the generator set within the same operating cycle to obtain the basic operating dataset;

[0044] The time window module is used to divide the basic running dataset into segments with the same time granularity, and to align and combine the data within each time granularity to obtain the time window running sequence.

[0045] The coupling module is used to calculate the coupling stability of the generator set's operating state within the operating cycle based on the time window operating sequence, and to obtain the coupling strength value;

[0046] The coordination module is used to group generator sets with coupling strength values ​​higher than a preset coupling value into the same coordination group, and arrange them into a coordination group sequence according to the change range of load response data within the coordination group.

[0047] The evolution module is used to calculate the degree of operational evolution stability of each cooperating group within the operating cycle based on the cooperating group sequence, and obtain the evolution stability value;

[0048] The scheduling module is used to assign scheduling weights to each coordination group based on the evolution stability value, and to map the scheduling weights to the unit start-up and shutdown priorities and load allocation ratios to obtain a coordination scheduling scheme.

[0049] The instruction module is used to generate operation and maintenance decision instructions based on the collaborative scheduling scheme, and to perform collaborative operation and maintenance on multiple generator sets according to the operation and maintenance decision instructions, so as to obtain the collaborative operation and maintenance results.

[0050] The above-described solution of the present invention has at least the following beneficial effects:

[0051] This invention divides the basic operational dataset into segments with the same time granularity and aligns and combines different types of data within each time granularity. This transforms the non-uniformly distributed time stream of continuous operational data into a time window sequence with clear boundaries and sequential relationships. In the time dimension, it eliminates the alignment deviation caused by inconsistent sampling frequencies of different monitoring data. This allows data within the same time granularity to be regarded as an objective representation of the same operational state. At the data processing level, it allows for direct comparison, accumulation, and sorting of the operational states of different time windows, providing an operable data format for subsequent management and scheduling analysis based on time evolution characteristics.

[0052] This invention calculates the coupling stability of generator unit operating states within an operating cycle using a time-window operating sequence. This transforms the implicit state of the relationship between units into an explicit numerical expression at the data level. The coupling strength value originates from the comprehensive calculation results of operating state changes within multiple time granularities, rather than a single instantaneous state judgment. In terms of data representation, it reflects the continuous characteristics of the unit operating relationship. It enables horizontal comparison and screening of the operating relationships between different units at the data level without relying on human experience or pre-set logical relationships. This provides a unified quantitative basis for identifying unit collaborative relationships, offering directly referable data indicators for subsequent collaborative grouping and scheduling management. It also forms an automatic judgment mechanism based on the evolution of operating data relationships in the management decision-making process.

[0053] This invention groups generator sets with coupling strength values ​​higher than a preset coupling value into the same cooperative group. This splits the multi-unit operation system, which originally required overall processing, into several cooperative units in terms of data structure. This splitting process is based on the calculation result of the coupling strength operation data rather than fixed grouping rules or manual configuration. This makes the formation of cooperative units have clear data basis and repeatability. The generator sets in each cooperative group have relatively consistent evolutionary characteristics in terms of changes in operating status. This allows subsequent scheduling and management to be carried out with the cooperative group as the basic processing object, without having to consider the complex relationships between all generator sets at the same time on a global scale. This achieves a structured decomposition of the operating object and helps to reduce the combinatorial complexity in subsequent scheduling calculations.

[0054] This invention calculates the degree of operational evolution stability of each collaborative group within its operating cycle to form an evolution stability value. This enables the collaborative groups to have a unified and comparable state evaluation index at the data level, which can reflect the continuous characteristics of the overall operating state of the collaborative group changing over time. By using the collaborative group as the evaluation object, the system can make decisions directly based on the group-level state in subsequent scheduling and management processes, without relying on the instantaneous performance of individual units. At the data processing level, it realizes the periodic summary and abstract expression of the operating state, providing a unified dimension for state comparison between different collaborative groups and between different operating cycles.

[0055] This invention maps evolutionary stability values ​​to scheduling weights and further to unit start-up / shutdown priorities and load allocation ratios, enabling the results of preceding data analysis to be directly transformed into executable operation and maintenance decision parameters at the data level. This provides a clear data source path for the generation of scheduling instructions, forming a complete data processing chain from operational data collection, status analysis, collaborative grouping to scheduling parameter output. The resulting operation and maintenance decision instructions do not rely on manual rule intervention but are directly driven by data calculation results, making scheduling behavior traceable and reproducible at the data level, and making the management and scheduling decision-making process a natural result of operational data processing. Attached Figure Description

[0056] Figure 1 This is a flowchart of a collaborative operation and maintenance decision-making method for multiple generator sets provided in an embodiment of the present invention. Detailed Implementation

[0057] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0058] like Figure 1 As shown, embodiments of the present invention propose a collaborative operation and maintenance decision-making method for multiple generator sets, the method comprising:

[0059] Acquire power change data, temperature change data, and load response data of the generator set within the same operating cycle to obtain a basic operating dataset;

[0060] The basic running dataset is divided into segments with the same time granularity, and the data within each time granularity is aligned and combined to obtain the time window running sequence;

[0061] Based on the time window operation sequence, the coupling stability of the generator set's operating state within the operating cycle is calculated to obtain the coupling strength value;

[0062] Generator sets with coupling strength values ​​higher than the preset coupling value are grouped into the same coordination group, and within the coordination group, they are arranged into a coordination group sequence according to the change range of load response data.

[0063] Based on the sequence of coordinating groups, the degree of operational evolution stability of each coordinating group within the operating cycle is calculated to obtain the evolution stability value;

[0064] Based on the evolutionary stability value, each coordination group is assigned a scheduling weight, and the scheduling weight is mapped to the unit start-up and shutdown priority and load allocation ratio to obtain the coordination scheduling scheme.

[0065] The operation and maintenance decision instructions are generated according to the collaborative scheduling scheme, and the collaborative operation and maintenance of multiple generator sets are carried out according to the operation and maintenance decision instructions to obtain the collaborative operation and maintenance results.

[0066] In this embodiment of the invention, power change data, temperature change data, and load response data of generator sets within the same operating cycle are acquired to obtain a basic operating dataset. This avoids the scattered storage or cross-cycle referencing of data from different operating states, providing consistent data input conditions for subsequent analysis. The basic operating dataset is divided into segments with the same time granularity, and the data within each time granularity is aligned and combined to obtain a time-window operating sequence. The operating states within different time granularities are processed window by window and compared across windows, providing a standardized data structure for subsequent calculations. Based on the time-window operating sequence, the coupling stability of the generator set's operating state within the operating cycle is calculated to obtain a coupling strength value. Units with cooperative characteristics in their operating state changes are identified, providing a clear data basis for subsequent cooperative grouping. Generator sets with coupling strength values ​​higher than a preset coupling value are grouped into the same cooperative group, and within the cooperative group, they are arranged into a cooperative group sequence according to the magnitude of load response data changes. This provides a clear processing object and order basis for subsequent state analysis and scheduling decisions for cooperative units.

[0067] Based on the coordination group sequence, the operational evolution stability of each coordination group within the operating cycle is calculated to obtain an evolution stability value. This provides a unified-dimensional group-level state evaluation result for subsequent scheduling decisions, avoiding the problem of making decisions based solely on a single point in time. Based on the evolution stability value, scheduling weights are assigned to each coordination group, and these weights are mapped to unit start-up / shutdown priorities and load allocation ratios to obtain a coordinated scheduling scheme. This ensures a clear data source path for the scheduling decision-making process, achieving a direct mapping from operational status data to scheduling strategies at the management decision-making level. Operation and maintenance decision instructions are generated based on the coordinated scheduling scheme, and coordinated operation and maintenance of multiple generating units is performed according to these instructions to obtain coordinated operation and maintenance results. This ensures the traceability of scheduling instruction generation, execution, and result recording, providing a reusable data foundation for subsequent operational analysis and operation and maintenance management.

[0068] This involves acquiring power variation data, temperature variation data, and load response data of the generator set within the same operating cycle to obtain a basic operating dataset, which specifically includes:

[0069] First, the operational cycle requiring collaborative operation and maintenance analysis is determined, and this cycle is used as the unified time range for this data collection and processing. Within this operational cycle, the system establishes a corresponding data acquisition channel for each generator unit participating in collaborative operation and maintenance. This channel communicates with the unit operation monitoring unit via a data interface to continuously receive raw monitoring data reflecting changes in the unit's operating status. This raw monitoring data includes at least power change data characterizing changes in unit output, temperature change data characterizing changes in the unit's thermal state, and load response data characterizing the unit's response to changes in external load. All of these data types carry corresponding unit identifiers and time identifiers during acquisition for subsequent data association and processing.

[0070] During data acquisition, the system performs preliminary analysis on the received raw monitoring data. Power change data is converted into a numerical form reflecting the output trend of the unit between adjacent sampling times; temperature change data is converted into a numerical form reflecting the thermal state of the unit over time; and load response data is converted into a numerical form reflecting the output adjustment behavior of the unit under load changes. This analysis process ensures that data from different sources and with different physical meanings have a unified numerical representation, facilitating joint processing and calculation in subsequent steps. Subsequently, the system filters the analyzed data based on the time boundaries of the operating cycle, retaining only data records whose time markers fall within the current operating cycle. The data is then aggregated according to the unit identifier, forming a set of operating data for each generator unit within that operating cycle, including power change data, temperature change data, and load response data. During this process, the system does not perform cross-cycle splicing or historical compensation; instead, it constructs data solely based on data collected within the current operating cycle, ensuring that the data used for subsequent analysis has a clear time range and a consistent operating context. After completing the data collection, the system will uniformly encapsulate the set of operating data generated by each generator set within the same operating cycle to construct a basic operating dataset.

[0071] In a preferred embodiment of the present invention, the basic running dataset is divided into segments with the same time granularity, and the data within each time granularity are aligned and combined to obtain a time window running sequence, including:

[0072] Based on the basic operation dataset, extract the time stamps for each power change data, temperature change data, and load response data, and organize the time stamps in the order of the operation cycle to obtain a time index set;

[0073] The time index set is divided into multiple consecutive time periods according to the same time granularity, and a segment number identifier is generated for each consecutive time period to obtain a time segmentation table.

[0074] Based on the time segmentation table, each data item in the basic operation dataset is assigned to the corresponding segment number according to its time identifier, and each data item is aggregated according to the unit number to obtain segmented aggregated data;

[0075] Based on the segmented aggregated data, each data item under the same segment number is matched according to the unit number, and the matching results are encapsulated into time window units to obtain time window combined data;

[0076] Arrange the time window combination data in ascending order according to the segment number identifier, and add a sequence position mark to each time window unit to obtain the time window running sequence.

[0077] In this embodiment of the invention, based on the basic operational dataset, time identifiers are extracted for each power change data, temperature change data, and load response data. These time identifiers are then organized according to the operational cycle sequence to obtain a time index set, avoiding alignment issues caused by scattered time identifiers across different types of operational data. The time index set is divided into multiple consecutive time periods with the same time granularity, and segment number identifiers are generated for each consecutive time period to obtain a time segmentation table. This ensures that there is no overlap or omission between different time periods, providing a standardized time framework for the segmentation and statistical processing of operational data. Based on the time segmentation table, each data item in the basic operational dataset is assigned according to its time identifier. The data is then aggregated by unit number to obtain segmented aggregated data. This allows for the analysis of the complete operating status of a single unit within a defined time period, avoiding data mixing issues across time periods or units. Based on the segmented aggregated data, each data point under the same segment number is matched by unit number, and the matching results are encapsulated into time window units to obtain time window combination data. This ensures that each time window unit can fully represent the operating status of a unit within a specific time period. The time window combination data is then sorted in ascending order by segment number and a sequence position mark is added to each time window unit to obtain the time window operating sequence, providing clear and complete data input units for subsequent calculations.

[0078] Specifically, based on the time segmentation table, each data entry in the basic operation dataset is assigned to its corresponding segment number according to its time identifier, and each data entry is aggregated by unit number to obtain segmented aggregated data, which specifically includes:

[0079] For any power change, temperature change, or load response data in the basic operation dataset, the system first reads the time identifier information carried in the data and compares it with the start and end time ranges corresponding to each segment identifier in the time segmentation table to determine the time period to which the data belongs. When the time identifier falls within the time range corresponding to a certain segment identifier, the system writes the segment identifier into the segment number field of the data, thus completing the time period assignment operation for the data. Next, the system reads the corresponding generator set number in the data and uses the segment identifier and generator set number as a combined index condition to aggregate the data that has been assigned time. Specifically, under the same segment identifier, data with the same generator set number are grouped into the same aggregation unit, and the power change data, temperature change data, and load response data contained in this aggregation unit are stored in a structured form. Through the above processing, the data in the basic operation dataset, which originally existed as single records, is reorganized into a data set with the segment identifier and generator set number as the basic index, forming segmented aggregated data.

[0080] Specifically, based on the segmented aggregated data, each data entry under the same segment number is matched according to the unit number, and the matching results are encapsulated into time window units to obtain time window combined data, which specifically includes:

[0081] For any given segment identifier, the system first extracts all aggregated data sets under that segment identifier and then performs matching and verification on these data sets based on the generator unit number to confirm whether the same generator unit number simultaneously contains power change data, temperature change data, and load response data. When it is confirmed that a generator unit number has complete data types under that segment identifier, the system encapsulates all data types corresponding to that generator unit number as a whole. The power change data, temperature change data, and load response data are written into the same data structure, and a unique time window identifier is assigned to this data structure to represent the corresponding time period and generator unit. This time window identifier is associated with the segment identifier and the generator unit number, enabling subsequent processing to directly locate the corresponding time period and generator unit through the time window identifier. The system sequentially performs the above matching and encapsulation operations on each generator unit number under the same segment identifier, forming multiple time window units within that time period. After completing the encapsulation of time window units under all segment identifiers, the system aggregates the time window units generated within each time period to form a combined time window data consisting of multiple time window units.

[0082] In a preferred embodiment of the present invention, the coupling stability of the generator set's operating state within the operating cycle is calculated based on the time window operating sequence to obtain the coupling strength value, including:

[0083] Based on the time window operation sequence, calculate whether the power change direction and temperature change direction are consistent within each time granularity to obtain the direction consistency suppression term; calculate the interaction value between the power change amplitude and temperature change amplitude within each time granularity to obtain the amplitude interaction contribution term; calculate the degree of disturbance of the load response amplitude within each time granularity to obtain the load disturbance reduction term.

[0084] Based on the direction consistency suppression term, the continuity of the direction consistency state within two adjacent time granularities is calculated to obtain the adjacent continuous enhancement term; based on the direction consistency suppression term, amplitude interaction contribution term, and load disturbance reduction term, the coupling contribution value within each time granularity is calculated to obtain the coupling contribution merging term.

[0085] Based on the coupling contribution merging term and adjacent consecutive enhancement terms, the enhanced coupling contribution value within each time granularity is calculated to obtain the coupling strength value.

[0086] In this embodiment of the invention, based on the time window operation sequence, the consistency between the power change direction and the temperature change direction within each time granularity is calculated to obtain a direction consistency suppression term, which weakens the impact of absolute value fluctuations on state judgment and identifies unit behavior patterns where the operating state direction is stable or frequently reverses; the interaction value between the power change amplitude and the temperature change amplitude within each time granularity is calculated to obtain an amplitude interaction contribution term, distinguishing between operating situations where the change direction is consistent but the amplitude is weakly correlated and those where the change direction is consistent and the amplitude is highly correlated, providing a quantifiable intensity reference for refined analysis in management and scheduling; the degree of disturbance of the load response amplitude within each time granularity is calculated to obtain a load disturbance reduction term, avoiding misjudging short-term synchronous changes under strong external disturbances as stable operating coupling relationships; based on the direction consistency suppression term, the following is calculated: The continuity of directional consistency within two adjacent time granularities is used to obtain adjacent continuous enhancement terms, distinguishing between occasional consistent behavior within isolated time granularities and consistent behavior that persists across multiple consecutive time granularities. Based on directional consistency suppression terms, amplitude interaction contribution terms, and load disturbance reduction terms, the coupling contribution value within each time granularity is calculated to obtain a coupling contribution merging term, providing standardized input for subsequent introduction of time continuity enhancement and periodic summarization, thus giving the data processing flow a clear hierarchical structure in management and scheduling analysis. Based on the coupling contribution merging term and adjacent continuous enhancement terms, the enhanced coupling contribution value within each time granularity is calculated to obtain a coupling strength value, enabling horizontal comparison and screening of the collaborative relationships between different units, providing directly usable data for subsequent collaborative grouping and scheduling strategy formulation.

[0087] In a preferred embodiment of the present invention, generator sets with coupling strength values ​​higher than a preset coupling value are grouped into the same cooperative group, and within the cooperative group, they are arranged into a cooperative group sequence according to the variation amplitude of load response data, including:

[0088] Generator sets with coupling strength values ​​higher than a preset coupling value are marked as candidate generator sets, and an index is established between the generator set number and the coupling strength value of the candidate generator sets to obtain a set of candidate generator sets;

[0089] A merging rule table is generated based on the index association of the candidate unit set, and the candidate units are assigned to the cooperative group identifier according to the merging rule table. When the same candidate unit satisfies multiple cooperative group identifiers at the same time, the candidate unit is merged into the cooperative group identifier with the highest coupling strength value to obtain the cooperative group set.

[0090] Based on the set of collaborative groups, the load response data of each generator set in the collaborative group during the operating cycle is extracted, and the peak-valley difference of the load response data and the cumulative value of the granularity change of adjacent time are combined into the load amplitude characteristic value to obtain the load amplitude value within the group.

[0091] Based on the load amplitude value within the group, the generator sets within the coordinated group are sorted according to the magnitude of the load amplitude characteristic value to obtain the coordinated group sequence.

[0092] In this embodiment of the invention, generator sets with coupling strength values ​​higher than a preset coupling value are marked as candidate generator sets. An index is established between the generator set number and the coupling strength value of each candidate generator set to obtain a candidate generator set. This avoids introducing weakly or unrelated generator sets into the collaborative modeling process, ensuring that the collaborative group formation process has a clear data source and traceability. Horizontal comparison and screening of collaborative relationships between different generator sets are performed. A merging rule table is generated based on the index association of the candidate generator set, and candidate generator sets are assigned to collaborative group identifiers according to the merging rule table. When the same candidate generator set simultaneously meets multiple collaborative group identifiers, the candidate generator set is merged into the collaborative group identifier with the highest coupling strength value, resulting in a collaborative group set. This ensures that the collaborative group set is within the collaborative group modeling process. The data layer possesses clear boundary conditions and uniqueness, providing structured input for subsequent independent operation and maintenance decisions for the collaborative group. Based on the collaborative group set, load response data of each generator unit within the collaborative group is extracted during its operating cycle. The peak-to-valley difference of the load response data and the cumulative value of changes in adjacent time granularities are combined to obtain the load amplitude feature value, thus obtaining the load amplitude value within the group. This avoids repeatedly traversing the original time series data during the sorting stage, forming a stable and reusable feature expression at the data processing level. Based on the load amplitude value within the group, the generator units within the collaborative group are sorted according to the magnitude of the load amplitude feature value to obtain the collaborative group sequence, providing a clear data reference path for subsequent scheduling weight mapping and start-stop sequence determination.

[0093] Specifically, a merging rule table is generated based on the index association of the candidate unit set, and the candidate units are assigned to the cooperative group identifier according to the merging rule table. When the same candidate unit satisfies multiple cooperative group identifiers at the same time, the candidate unit is merged into the cooperative group identifier with the highest coupling strength value to obtain the cooperative group set, which specifically includes:

[0094] First, the system reads the generator unit numbers corresponding to each generator unit in the candidate generator set and the coupling strength values ​​associated with each generator unit number. Using the generator unit number as the primary index field and the coupling strength value as the association judgment field, a merging rule table is constructed to describe the merging relationships between candidate generator sets. Each rule in the merging rule table characterizes the association conditions of candidate generator sets in the coupling strength dimension, and by setting a range of coupling strength values ​​or relative magnitudes, the system limits the range of cooperative group identifiers to which candidate generator sets can be assigned. During the merging operation, the system matches and judges candidate generator sets one by one according to the merging rule table, and assigns those that meet the criteria... Candidate units with the same merging condition are temporarily mapped to their corresponding cooperative group identifiers. When a candidate unit is detected to simultaneously meet the merging conditions corresponding to multiple cooperative group identifiers, the system reads the coupling strength value of the candidate unit under different merging conditions, compares the coupling strength values, merges the candidate unit into the cooperative group identifier with the largest coupling strength value, and cancels its affiliation under the other cooperative group identifiers. Through the above processing, each candidate unit belongs to only one cooperative group identifier in the data structure, ultimately forming a cooperative group set consisting of multiple cooperative group identifiers and their corresponding unit sets.

[0095] Specifically, based on the set of collaborative groups, the load response data of each generator set in the collaborative group during the operating cycle is extracted, and the peak-valley difference of the load response data and the cumulative value of the granular change of adjacent time are combined into the load amplitude characteristic value to obtain the load amplitude value within the group, which specifically includes:

[0096] The system uses the collaborative group identifier as an index to perform data processing operations within each collaborative group. For any collaborative group, it first determines the set of generator sets belonging to that group based on the collaborative group identifier, and then extracts the load response data sequence corresponding to the generator sets within the same operating cycle. During the load response data processing, the system traverses the load response data sequence in time granularity order, determines the maximum and minimum values ​​of the load response data within the operating cycle, and calculates the difference between the maximum and minimum values ​​as the peak-valley difference of the load response. Simultaneously, the system processes data in order of adjacent time granularities. The system sequentially calculates the changes in load response data at adjacent time granularities and accumulates these changes to obtain a cumulative value representing the overall fluctuation process of the load response. After calculating the peak-valley difference and the cumulative value of changes, the system combines these values ​​to generate a load amplitude characteristic value that describes the load response behavior of the generator set during its operating cycle. This load amplitude characteristic value is then associated with and stored with the corresponding generator set number. Through this process, each generator set within the collaborative group receives a load amplitude characteristic value that comprehensively reflects its load response range and change process.

[0097] In a preferred embodiment of the present invention, the degree of operational evolution stability of each cooperative group within the operating cycle is calculated based on the cooperative group sequence to obtain an evolution stability value, including:

[0098] Based on the cooperative group sequence, the mean values ​​of power change data and temperature change data of each generator set in the cooperative group within each time granularity are calculated and used as power reference trajectory and temperature reference trajectory to obtain the reference trajectory term within the group; the deviation of power change data of each generator set in the cooperative group from the power reference trajectory within each time granularity is calculated to obtain the dispersion term within the group.

[0099] Based on the intra-group reference trajectory term, the continuity of the evolution trajectory of the power reference trajectory and temperature reference trajectory between adjacent time granularities is calculated to obtain the intra-group smoothness term; the aggregation amplitude of the load response data of each unit in the cooperative group within each time granularity is calculated to obtain the disturbance gating term.

[0100] Based on the intra-group dispersion term, intra-group smoothness term, and disturbance gating term, the stable contribution value of each time granularity is calculated to obtain the single-time-window stable contribution term; based on the single-time-window stable contribution term and the disturbance gating term, the cumulative stable contribution value of all time granularities within the running cycle is calculated to obtain the evolutionary stable value.

[0101] In this embodiment of the invention, based on the coordinated group sequence, the mean values ​​of power change data and temperature change data of each generator unit in the coordinated group within each time granularity are calculated and used as power reference trajectories and temperature reference trajectories, respectively, to obtain the intra-group reference trajectory term. This eliminates the offset influence caused by instantaneous outliers of individual units and avoids repeatedly switching reference benchmarks between multiple unit data in subsequent analysis. The deviation between the power change data of each generator unit in the coordinated group and the power reference trajectory within each time granularity is calculated to obtain the intra-group dispersion term, which characterizes the degree of dispersion of the operating status of each unit within the coordinated group relative to the overall trend of the group. Based on the intra-group reference trajectory term, the continuity of the evolution trajectory of the power reference trajectory and temperature reference trajectory between adjacent time granularities is calculated to obtain the intra-group smoothness term. This quantitatively expresses the evolution continuity of the operating status of the coordinated group from the time dimension, reflecting the change of the reference trajectory in the time series. Whether the evolution exhibits continuous and gradual characteristics; calculate the aggregation amplitude of the load response data of each unit in the collaborative group within each time granularity to obtain the disturbance gating term, avoiding the mixing of state fluctuations caused by drastic load changes with internal operational anomalies, and enabling the operational data to have conditional differentiation capabilities in management analysis; calculate the stable contribution value of each time granularity based on the intra-group dispersion term, intra-group smoothness term, and disturbance gating term to obtain the single-time-window stable contribution term, reflecting the internal consistency state, time evolution pattern, and external load environment conditions of the collaborative group at that time granularity, forming a comprehensive description of the operating state of a single time slice; calculate the cumulative stable contribution value of all time granularities within the operating cycle based on the single-time-window stable contribution term and disturbance gating term, obtaining the evolutionary stable value, reflecting the comprehensive performance of the state of multiple time slices within the operating cycle, and used for state comparison between different collaborative groups or different operating cycles.

[0102] In a preferred embodiment of the present invention, scheduling weights are assigned to each cooperative group based on the evolutionary stability value, and the scheduling weights are mapped to unit start-up / shutdown priorities and load allocation ratios to obtain a cooperative scheduling scheme, including:

[0103] The evolutionary stability value of each coordination group is mapped to a scheduling weight, and the scheduling weight is continuously corrected according to the preset weight jump restriction rule to obtain the coordination group scheduling weight.

[0104] The scheduling weight of the coordination group is converted into the start-stop priority of the coordination group, and the group order in the coordination group sequence is inherited as the start-stop order of the group. At the same time, the start-stop interval constraint table is generated according to the coupling strength value of the units in the coordination group, and the start-stop interval constraint table is written into the start-stop priority table to obtain the unit start-stop priority.

[0105] The load demand data within the operating cycle is divided into load segments according to time granularity, and the load segments are proportionally allocated according to the scheduling weight of the coordination group. A proportional adjustment threshold table is generated based on the evolution and stability value of each coordination group to correct the proportional allocation results within the same time granularity, thus obtaining the load allocation ratio.

[0106] In this embodiment of the invention, the evolutionary stability value of each coordination group is mapped to a scheduling weight, and the scheduling weight is continuously corrected according to a preset weight jump restriction rule to obtain the coordination group scheduling weight. This reflects the relative position of the coordination group in the scheduling system and avoids discontinuous changes in scheduling parameters over time, reflecting the comprehensive performance of multiple time slice states within the operating cycle. The coordination group scheduling weight is converted into the coordination group start-up and shutdown priority, and the intra-group sorting in the coordination group sequence is inherited as the intra-group start-up and shutdown order. At the same time, a start-up and shutdown interval constraint table is generated based on the coupling strength value of the units within the coordination group, and the start-up and shutdown interval constraint table is written into the start-up and shutdown priority table to obtain the unit start-up and shutdown priority. Firstly, the coupling relationship between units is explicitly transformed into start-stop interval constraints, providing clear and directly analyzable start-stop decision-making basis for subsequent scheduling execution. Secondly, load demand data within the operating cycle is divided into load segments according to time granularity, and these load segments are proportionally allocated according to the scheduling weight of the coordination group. Thirdly, a proportional adjustment threshold table is generated based on the evolutionary stability value of each coordination group to correct the proportional allocation results within the same time granularity, resulting in a load allocation ratio. This decomposes the load demand data into independently processable load units in the time dimension, forming a stable and controllable load allocation ratio structure, providing clear load configuration parameters for scheduling execution.

[0107] Specifically, the evolutionary stability value of each coordination group is mapped to a scheduling weight, and the scheduling weight is continuously corrected according to a preset weight jump restriction rule to obtain the coordination group scheduling weight, which includes:

[0108] After a system cycle ends, it first reads the evolutionary stable values ​​corresponding to each cooperative group within that cycle from the storage unit and organizes them into a cooperative group-stable value mapping table using the cooperative group identifier as an index. Then, the system pre-sets a mapping interval table for weight generation in the scheduling module. This mapping interval table maps the numerical range of the evolutionary stable values ​​to the value range of the scheduling weights. For example, the evolutionary stable values ​​are divided into several continuous intervals, and a corresponding weight calculation coefficient is preset for each interval. The system performs interval matching on the evolutionary stable values ​​of each cooperative group, determines its corresponding mapping interval, and then performs numerical conversion on the evolutionary stable values ​​based on the weight calculation coefficient corresponding to that mapping interval to generate the initial scheduling weight for that cooperative group in the current cycle. It further reads the historical scheduling weights that were effective in the previous cycle for that cooperative group and calculates the difference between the initial scheduling weight generated in the current cycle and the historical scheduling weights to obtain the weight change. The system then calls a preset weight jump restriction rule to judge the weight change, whereby the weight jump restriction rule limits the maximum allowable change in the scheduling weight of the same cooperative group within adjacent cycles. When a weight change exceeds the maximum allowable range, the system corrects the initial scheduling weight for the current period, making its value equal to the sum or difference of the historical scheduling weight and the maximum allowable change range, thus limiting the range of scheduling weight changes over time. After this processing, the system uses the corrected scheduling weight as the final effective scheduling weight for the collaborative group in the current operating period and writes it into the scheduling weight table.

[0109] Specifically, the scheduling weights of the coordination group are converted into coordination group start-up and shutdown priorities, and the intra-group sorting in the coordination group sequence is inherited as the intra-group start-up and shutdown order. Simultaneously, a start-up and shutdown interval constraint table is generated based on the coupling strength values ​​of the units within the coordination group, and this table is written into the start-up and shutdown priority table to obtain the unit start-up and shutdown priorities. This includes:

[0110] First, the scheduling weights of all coordinated groups are sorted based on their numerical values. The system generates a priority sequence for coordinated groups according to their weights, either from high to low or low to high, and converts this sequence into a coordinated group start-stop priority identifier, giving each coordinated group a unique and definite priority number in start-stop scheduling. Then, for each coordinated group, the system reads the coordinated group sequence formed in the previous processing. This sequence records the ranking relationship of each generator unit within the coordinated group based on load response changes. The system directly maps this ranking relationship to the start-stop order within the group, setting the units ranked higher in the sequence as priority start-stop units and the units ranked lower as subsequent start-stop units, forming a start-stop order table within the group. The system further obtains the coupling strength value corresponding to each generator unit within the coordinated group and compares the coupling strength of any two units within the coordinated group by pairwise combinations of their unit numbers. The system generates a start-stop interval constraint table based on the comparison results. When the coupling strength between two units is higher than a preset coupling threshold, the system sets a larger start-stop interval constraint for that unit; conversely, when the coupling strength is lower than the threshold, a smaller start-stop interval constraint is set. The start-stop interval constraint is represented by a time length or the number of time slices. Finally, the system associates and stores the start-stop interval constraint table with the cooperative group start-stop priority identifier and the group's start-stop sequence table to form a unit start-stop priority table.

[0111] Specifically, the load demand data within the operating cycle is divided into load segments according to time granularity, and the load segments are proportionally allocated according to the scheduling weight of the coordination group. A proportional adjustment threshold table is generated based on the evolutionary stability value of each coordination group to correct the proportional allocation results within the same time granularity, thus obtaining the load allocation ratio. This includes:

[0112] First, the system retrieves load demand data corresponding to the operating cycle from the load monitoring module. Following a time granularity consistent with the operating data processing phase, the continuous load demand data is segmented, representing the load demand within the operating cycle as multiple load segments arranged chronologically. For each load segment corresponding to a time granularity, the system treats it as an independent load allocation object. Subsequently, at the same time granularity, the system reads the scheduling weights of all collaborative groups and calculates the proportion of each collaborative group's scheduling weight in the total weight. Based on this proportion, the system allocates the load segments corresponding to the current time granularity proportionally to each collaborative group, giving each collaborative group an initial load allocation ratio corresponding to its scheduling weight proportion. The system further reads the evolutionary stability value corresponding to each collaborative group and generates a proportional adjustment threshold table based on the evolutionary stability value. This proportional adjustment threshold table limits the upper and lower limits of the allowable load proportion that each collaborative group can undertake within a single time granularity. The system compares the initial load allocation ratio with the ratio adjustment threshold table one by one. When it detects that the load allocation ratio of a certain collaborative group exceeds its corresponding upper limit threshold or falls below its lower limit threshold, the system adjusts the load allocation ratio to bring it within the corresponding threshold range. Simultaneously, the system performs linked corrections on the load allocation ratios of other collaborative groups to ensure that the sum of the load allocation ratios of all collaborative groups remains constant within the same time granularity. After the above processing, the system outputs the final effective load allocation ratio at that time granularity and writes the load allocation ratio results corresponding to each time granularity into the load allocation table.

[0113] In a preferred embodiment of the present invention, an operation and maintenance decision instruction is generated according to the collaborative scheduling scheme, and collaborative operation and maintenance is performed on multiple generator sets according to the operation and maintenance decision instruction to obtain a collaborative operation and maintenance result, including:

[0114] Based on the collaborative scheduling scheme, the unit start-up and shutdown priority, start-up and shutdown interval constraint table and load allocation ratio are extracted, and a parameter mapping relationship is established between them and the unit identifier according to the time granularity identifier to obtain the instruction parameter set.

[0115] The instruction parameter set is split into start / stop instruction fragments and load instruction fragments, and an execution order identifier is written for each instruction fragment to obtain the instruction fragment set;

[0116] Based on the instruction fragment set, conflict screening is performed on start / stop instruction fragments and load instruction fragments under the same time granularity identifier. The conflicting instruction fragments are then adjusted according to the start / stop interval constraint table and the total constraint of load allocation ratio to obtain the operation and maintenance decision instruction set.

[0117] Based on the operation and maintenance decision instruction set, the unit start-up and shutdown control instructions and load distribution control instructions are output sequentially, and the execution status of each time granularity is written into the operation and maintenance record table to obtain the collaborative operation and maintenance results.

[0118] In this embodiment of the invention, based on the collaborative scheduling scheme, the unit start-stop priority, start-stop interval constraint table, and load allocation ratio are extracted, and a parameter mapping relationship is established between them and the unit identifier according to the time granularity identifier to obtain the instruction parameter set. This avoids re-parsed scheduling logic during the execution phase, achieving a clear implementation of scheduling results into execution parameters. The instruction parameter set is split into start-stop instruction fragments and load instruction fragments, and an execution order identifier is written for each instruction fragment to obtain an instruction fragment set. This makes start-stop control and load control independent in data structure, facilitating subsequent separate verification and scheduling processing of different types of instructions. Based on the instruction fragment set, in the same... Under time granularity identification, conflict screening is performed on start / stop command fragments and load command fragments. Conflicting command fragments are adjusted according to the start / stop interval constraint table and the total constraint of load allocation ratio to obtain the operation and maintenance decision command set. This ensures that the commands comply with scheduling rules and operational constraints, avoiding command conflicts or constraint violations during the execution phase. Based on the operation and maintenance decision command set, unit start / stop control commands and load allocation control commands are output sequentially, and the execution status of each time granularity is written into the operation and maintenance record table to obtain collaborative operation and maintenance results. The scheduling decision is gradually implemented into executable control behavior in the time dimension, forming a closed loop from decision generation to execution feedback.

[0119] Specifically, the instruction parameter set is split into start / stop instruction fragments and load instruction fragments, and an execution order identifier is written for each instruction fragment to obtain the instruction fragment set, which includes:

[0120] First, the system parses each instruction parameter set line by line, treating each instruction parameter record as a basic processing object. For each instruction parameter record, the system performs functional discrimination based on the parameter types contained in the record. When the instruction parameter record contains parameters corresponding to the unit start / stop priority and start / stop interval constraint table, the system extracts the parameter information related to unit start / stop from the instruction parameter record and generates the corresponding start / stop instruction fragment using the unit identifier and time granularity identifier as association indexes. When the instruction parameter record contains a load allocation ratio, the system extracts the load allocation ratio and its corresponding time granularity identifier and unit identifier from the instruction parameter record to generate the corresponding load instruction fragment. Subsequently, based on the collaborative group scheduling weight and unit start / stop priority relationship determined in the collaborative scheduling scheme, the system sequentially marks the generated start / stop instruction fragments and load instruction fragments, writing the execution order identifier reflecting the execution sequence into each instruction fragment, so that each instruction fragment has a clear execution time and execution order attribute. Through the above processing, the scheduling results, originally existing in the form of parameters, are transformed into a set of instruction fragments with clear functions and a clear order, resulting in an instruction fragment set.

[0121] Specifically, based on the instruction fragment set, conflict screening is performed on start / stop instruction fragments and load instruction fragments under the same time granularity identifier. Then, the conflicting instruction fragments are adjusted according to the start / stop interval constraint table and the total constraint of load allocation ratio, resulting in an operation and maintenance decision instruction set, which specifically includes:

[0122] The system uses time granularity identifiers as a screening dimension to aggregate start / stop instruction fragments and load instruction fragments with the same time granularity identifier in the instruction fragment set, and constructs a corresponding subset of instructions to be executed within each time granularity. For the subset of instructions to be executed, the system first performs a consistency check on the start / stop instruction fragments to determine whether there are multiple start / stop instruction fragments corresponding to the same unit within the same time granularity, or whether start / stop instruction fragments within adjacent time granularities violate the minimum time interval requirement defined in the start / stop interval constraint table. When an interval conflict is detected between start / stop instruction fragments, the system adjusts or postpones the start / stop instruction fragments with lower priority or subsequent execution based on start / stop priority and execution order identifiers. At the same time, the system performs a total quantity check on the load instruction fragments within the same time granularity to determine whether the sum of the load allocation ratios corresponding to each load instruction fragment meets the total quantity constraint of the load allocation ratio. When the load allocation ratio exceeds or falls short of the preset total quantity constraint, the system adjusts the load allocation ratio of the relevant load instruction fragments according to the scheduling weight relationship of the coordination group. Subsequently, the adjusted start / stop instruction fragments and load instruction fragments were re-verified for consistency to confirm that they simultaneously met the start / stop interval constraint table and the total load allocation ratio constraint requirements. The verified instruction fragments were then re-aggregated according to their execution order to form a set of operation and maintenance decision instructions that can be directly executed at each time granularity.

[0123] Embodiments of the present invention also provide a collaborative operation and maintenance decision-making system for multiple generator sets, the system comprising:

[0124] The data module is used to acquire power change data, temperature change data, and load response data of the generator set within the same operating cycle to obtain the basic operating dataset;

[0125] The time window module is used to divide the basic running dataset into segments with the same time granularity, and to align and combine the data within each time granularity to obtain the time window running sequence.

[0126] The coupling module is used to calculate the coupling stability of the generator set's operating state within the operating cycle based on the time window operating sequence, and to obtain the coupling strength value;

[0127] The coordination module is used to group generator sets with coupling strength values ​​higher than a preset coupling value into the same coordination group, and arrange them into a coordination group sequence according to the change range of load response data within the coordination group.

[0128] The evolution module is used to calculate the degree of operational evolution stability of each cooperating group within the operating cycle based on the cooperating group sequence, and obtain the evolution stability value;

[0129] The scheduling module is used to assign scheduling weights to each coordination group based on the evolution stability value, and to map the scheduling weights to the unit start-up and shutdown priorities and load allocation ratios to obtain a coordination scheduling scheme.

[0130] The instruction module is used to generate operation and maintenance decision instructions based on the collaborative scheduling scheme, and to perform collaborative operation and maintenance on multiple generator sets according to the operation and maintenance decision instructions, so as to obtain the collaborative operation and maintenance results.

[0131] It should be noted that this system is a system corresponding to the above method. All implementation methods in the above method embodiments are applicable to this embodiment and can achieve the same technical effect.

[0132] Embodiments of the present invention also provide a computing device, including: a processor and a memory storing a computer program, wherein the computer program, when executed by the processor, performs the method described above. All implementations in the above method embodiments are applicable to this embodiment and can achieve the same technical effects.

[0133] Embodiments of the present invention also provide a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the method described above. All implementations in the above method embodiments are applicable to this embodiment and can achieve the same technical effects.

[0134] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A collaborative operation and maintenance decision-making method for multiple generator sets, characterized in that, The method includes: Acquire power change data, temperature change data, and load response data of the generator set within the same operating cycle to obtain a basic operating dataset; The basic running dataset is divided into segments with the same time granularity, and the data within each time granularity is aligned and combined to obtain the time window running sequence; Based on the time window operation sequence, calculate whether the power change direction and temperature change direction are consistent within each time granularity to obtain the direction consistency suppression term; calculate the interaction value between the power change amplitude and temperature change amplitude within each time granularity to obtain the amplitude interaction contribution term; calculate the degree of disturbance of the load response amplitude within each time granularity to obtain the load disturbance reduction term. Based on the direction consistency suppression term, the continuity of the direction consistency state within two adjacent time granularities is calculated to obtain the adjacent continuous enhancement term; based on the direction consistency suppression term, amplitude interaction contribution term, and load disturbance reduction term, the coupling contribution value within each time granularity is calculated to obtain the coupling contribution merging term. Based on the coupling contribution merging term and adjacent consecutive enhancement terms, the enhanced coupling contribution value within each time granularity is calculated to obtain the coupling strength value. Generator sets with coupling strength values ​​higher than the preset coupling value are grouped into the same coordination group, and within the coordination group, they are arranged into a coordination group sequence according to the change range of load response data. Based on the cooperative group sequence, the mean values ​​of power change data and temperature change data of each generator set in the cooperative group within each time granularity are calculated and used as power reference trajectory and temperature reference trajectory to obtain the reference trajectory term within the group; the deviation of power change data of each generator set in the cooperative group from the power reference trajectory within each time granularity is calculated to obtain the dispersion term within the group. Based on the intra-group reference trajectory term, the continuity of the evolution trajectory of the power reference trajectory and temperature reference trajectory between adjacent time granularities is calculated to obtain the intra-group smoothness term; the aggregation amplitude of the load response data of each unit in the cooperative group within each time granularity is calculated to obtain the disturbance gating term. Based on the intra-group dispersion term, intra-group smoothness term, and disturbance gating term, calculate the stable contribution value of each time granularity to obtain the single-time-window stable contribution term; based on the single-time-window stable contribution term and the disturbance gating term, calculate the cumulative stable contribution value of all time granularities within the running cycle to obtain the evolutionary stable value. Based on the evolutionary stability value, each coordination group is assigned a scheduling weight, and the scheduling weight is mapped to the unit start-up and shutdown priority and load allocation ratio to obtain the coordination scheduling scheme. The operation and maintenance decision instructions are generated according to the collaborative scheduling scheme, and the collaborative operation and maintenance of multiple generator sets are carried out according to the operation and maintenance decision instructions to obtain the collaborative operation and maintenance results.

2. The collaborative operation and maintenance decision-making method for multiple generator sets according to claim 1, characterized in that, The basic runtime dataset is divided into segments with the same time granularity, and the data within each time granularity is aligned and combined to obtain the time window runtime sequence, including: Based on the basic operation dataset, extract the time stamps for each power change data, temperature change data, and load response data, and organize the time stamps in the order of the operation cycle to obtain a time index set; The time index set is divided into multiple consecutive time periods according to the same time granularity, and a segment number identifier is generated for each consecutive time period to obtain a time segmentation table. Based on the time segmentation table, each data item in the basic operation dataset is assigned to the corresponding segment number according to its time identifier, and each data item is aggregated according to the unit number to obtain segmented aggregated data; Based on the segmented aggregated data, each data item under the same segment number is matched according to the unit number, and the matching results are encapsulated into time window units to obtain time window combined data; Arrange the time window combination data in ascending order according to the segment number identifier, and add a sequence position mark to each time window unit to obtain the time window running sequence.

3. The collaborative operation and maintenance decision-making method for multiple generator sets according to claim 2, characterized in that, Generator sets with coupling strength values ​​higher than a preset coupling value are grouped into the same coordination group, and within the coordination group, they are arranged into a coordination group sequence according to the magnitude of change in load response data, including: Generator sets with coupling strength values ​​higher than a preset coupling value are marked as candidate generator sets, and an index is established between the generator set number and the coupling strength value of the candidate generator sets to obtain a set of candidate generator sets; A merging rule table is generated based on the index association of the candidate unit set, and the candidate units are assigned to the cooperative group identifier according to the merging rule table. When the same candidate unit satisfies multiple cooperative group identifiers at the same time, the candidate unit is merged into the cooperative group identifier with the highest coupling strength value to obtain the cooperative group set. Based on the set of collaborative groups, the load response data of each generator set in the collaborative group during the operating cycle is extracted, and the peak-valley difference of the load response data and the cumulative value of the granularity change of adjacent time are combined into the load amplitude characteristic value to obtain the load amplitude value within the group. Based on the load amplitude value within the group, the generator sets within the coordinated group are sorted according to the magnitude of the load amplitude characteristic value to obtain the coordinated group sequence.

4. The collaborative operation and maintenance decision-making method for multiple generator sets according to claim 3, characterized in that, Based on the evolutionary stability value, scheduling weights are assigned to each coordination group, and these scheduling weights are mapped to unit start / stop priorities and load allocation ratios to obtain a coordinated scheduling scheme, including: The evolutionary stability value of each coordination group is mapped to a scheduling weight, and the scheduling weight is continuously corrected according to the preset weight jump restriction rule to obtain the coordination group scheduling weight. The scheduling weight of the coordination group is converted into the start-stop priority of the coordination group, and the group order in the coordination group sequence is inherited as the start-stop order of the group. At the same time, the start-stop interval constraint table is generated according to the coupling strength value of the units in the coordination group, and the start-stop interval constraint table is written into the start-stop priority table to obtain the unit start-stop priority. The load demand data within the operating cycle is divided into load segments according to time granularity, and the load segments are proportionally allocated according to the scheduling weight of the coordination group. A proportional adjustment threshold table is generated based on the evolution and stability value of each coordination group to correct the proportional allocation results within the same time granularity, thus obtaining the load allocation ratio.

5. The collaborative operation and maintenance decision-making method for multiple generator sets according to claim 4, characterized in that, Based on the collaborative scheduling scheme, operation and maintenance decision instructions are generated, and collaborative operation and maintenance of multiple generator units are performed according to the operation and maintenance decision instructions to obtain collaborative operation and maintenance results, including: Based on the collaborative scheduling scheme, the unit start-up and shutdown priority, start-up and shutdown interval constraint table and load allocation ratio are extracted, and a parameter mapping relationship is established between them and the unit identifier according to the time granularity identifier to obtain the instruction parameter set. The instruction parameter set is split into start / stop instruction fragments and load instruction fragments, and an execution order identifier is written for each instruction fragment to obtain the instruction fragment set; Based on the instruction fragment set, conflict screening is performed on start / stop instruction fragments and load instruction fragments under the same time granularity identifier. The conflicting instruction fragments are then adjusted according to the start / stop interval constraint table and the total constraint of load allocation ratio to obtain the operation and maintenance decision instruction set. Based on the operation and maintenance decision instruction set, the unit start-up and shutdown control instructions and load distribution control instructions are output sequentially, and the execution status of each time granularity is written into the operation and maintenance record table to obtain the collaborative operation and maintenance results.

6. A collaborative operation and maintenance decision-making system for multiple generator sets, characterized in that: The system is used to perform the method as described in any one of claims 1 to 5, the system comprising: The data module is used to acquire power change data, temperature change data, and load response data of the generator set within the same operating cycle to obtain the basic operating dataset; The time window module is used to divide the basic running dataset into segments with the same time granularity, and to align and combine the data within each time granularity to obtain the time window running sequence. The coupling module is used to calculate the coupling stability of the generator set's operating state within the operating cycle based on the time window operating sequence, and to obtain the coupling strength value; The coordination module is used to group generator sets with coupling strength values ​​higher than a preset coupling value into the same coordination group, and arrange them into a coordination group sequence according to the change range of load response data within the coordination group. The evolution module is used to calculate the degree of operational evolution stability of each cooperating group within the operating cycle based on the cooperating group sequence, and obtain the evolution stability value; The scheduling module is used to assign scheduling weights to each coordination group based on the evolution stability value, and to map the scheduling weights to the unit start-up and shutdown priorities and load allocation ratios to obtain a coordination scheduling scheme. The instruction module is used to generate operation and maintenance decision instructions based on the collaborative scheduling scheme, and to perform collaborative operation and maintenance on multiple generator sets according to the operation and maintenance decision instructions, so as to obtain the collaborative operation and maintenance results.

7. A computing device, characterized in that, include: One or more processors; A storage device for storing one or more programs, which, when executed by one or more processors, cause the one or more processors to implement the method as described in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a program that, when executed by a processor, implements the method as described in any one of claims 1 to 5.

Images