Campus comprehensive energy intelligent monitoring and optimal scheduling method and system

Abstract

The application discloses a park comprehensive energy intelligent monitoring and optimal scheduling method and system, collects energy consumption data and equipment operation state data, performs energy level margin gradient distribution according to equipment importance to construct a hierarchical load chain; in combination with the hierarchical load chain and the equipment operation state data, an energy consumption loss area is identified, a scheduling advantage area is formed based on recovery benefit evaluation; load density analysis is performed on the scheduling advantage area to form a density fluctuation curve to identify a transferable load set, load balance flow is generated by comprehensively responding to willingness and transfer capacity; scheduling redundant resources are extracted from the load balance flow, resource mutual exclusion relationship identification is obtained through loop correlation analysis, and adjustable load resource pools are formed according to response speed classification; supply and demand risk detection is performed on the load balance flow to determine an energy early warning level, a dynamic scheduling network is triggered by releasing priority to form resource pool construction; a partition scheduling configuration is generated by partitioning and scheduling the dynamic scheduling network, and a global coordination output scheduling control command is integrated.

Description

Technical Field

[0001] This invention relates to the field of energy management technology, and in particular to a method and system for intelligent monitoring and optimized scheduling of integrated energy resources in industrial parks. Background Technology

[0002] The park's integrated energy system integrates multiple energy forms such as electricity, natural gas, steam, and heating and cooling, improving overall energy efficiency through cascaded energy utilization and multi-energy complementarity. As the park expands and the number of energy-consuming devices increases, the complexity of energy dispatch and management continues to rise, and traditional manual dispatch methods are insufficient to meet the collaborative optimization needs of multiple energy sources, multiple devices, and multiple time periods.

[0003] The existing park energy management system has the following limitations in terms of scheduling optimization: First, the assessment of the scheduling potential of energy-consuming equipment is not precise enough, and it fails to distinguish the importance level of equipment and the user's willingness to respond, resulting in a lack of targeted selection of scheduling resources; second, there is a lack of analysis on the conflict relationship between scheduling resources, and the simultaneous call of resources on shared circuits may cause overload risks; third, the scheduling strategy mostly adopts a centralized architecture, and the scheduling response speed is affected when communication fails or the system load is too heavy. Summary of the Invention

[0004] This invention discloses a method and system for intelligent monitoring and optimized scheduling of integrated energy in industrial parks. It aims to construct a hierarchical load chain by classifying equipment importance, delineate scheduling advantage areas based on recovery benefits, generate load balancing flows by combining user response potential, form an adjustable load resource pool by calibrating resource mutual exclusion relationships through loop correlation analysis, and construct a dynamic scheduling network based on the resource pool according to the early warning level. This enables the integration of regional autonomy and global coordination to output scheduling control commands, providing a refined scheduling control scheme for integrated energy systems in industrial parks.

[0005] The first aspect of this invention proposes a method for intelligent monitoring and optimized scheduling of integrated energy resources in industrial parks, comprising the following steps:

[0006] Collect comprehensive energy consumption data and equipment operation status data of the park, and construct a hierarchical load chain based on the energy consumption data and the equipment operation status data to perform energy level cascade matching.

[0007] By combining the hierarchical load chain and the equipment operating status data, energy loss areas are identified, and the energy loss areas are classified according to their recoverability to form scheduling advantage areas.

[0008] Load density analysis is performed on the scheduling advantage area to form a density fluctuation curve. Based on the density fluctuation curve, a set of transferable loads is identified. Based on the set of transferable loads, user-side response potential is assessed to generate a load balancing flow.

[0009] The load balancing flow is subjected to scheduling capacity analysis to extract scheduling redundancy resources. The scheduling redundancy resources are subjected to call conflict analysis to obtain resource mutual exclusion relationship identifiers. Based on the resource mutual exclusion relationship identifiers, an adjustable load resource pool is formed according to the response speed.

[0010] For the load balancing flow, supply and demand risk detection is performed to obtain energy early warning level, and multi-level linkage is triggered according to the energy early warning level to generate a dynamic scheduling network from the adjustable load resource pool.

[0011] Based on the dynamic scheduling network, partition scheduling is performed to generate partition scheduling configurations, and the partition scheduling configurations are globally coordinated and integrated to output scheduling control commands.

[0012] The second aspect of this invention proposes a smart monitoring and optimized scheduling system for integrated energy in a park, comprising:

[0013] The data acquisition module is used to collect comprehensive energy consumption data and equipment operation status data in the park, and to construct a hierarchical load chain based on the energy consumption data and the equipment operation status data for energy level matching.

[0014] The loss identification module is used to identify energy loss areas by combining the hierarchical load chain and the equipment operating status data, and to classify the energy loss areas according to the loss recoverability to form scheduling advantage areas.

[0015] The load analysis module is used to perform load density analysis on the scheduling advantage area to form a density fluctuation curve, identify a set of transferable loads based on the density fluctuation curve, and perform user-side response potential assessment based on the set of transferable loads to generate a load balancing flow.

[0016] The resource management module is used to perform scheduling capacity analysis on the load balancing flow to extract scheduling redundant resources, perform call conflict analysis on the scheduling redundant resources to obtain resource mutual exclusion relationship identifiers, and form an adjustable load resource pool based on the resource mutual exclusion relationship identifiers according to the response speed.

[0017] The early warning response module is used to detect supply and demand risks for the load balancing flow, obtain energy early warning levels, and trigger the adjustable load resource pool to generate a dynamic scheduling network according to the energy early warning levels.

[0018] The coordination and scheduling module is used to generate partitioned scheduling configurations based on the dynamic scheduling network, and to globally coordinate and integrate the partitioned scheduling configurations to output scheduling control commands.

[0019] The beneficial effects of this invention are reflected in the following points: 1. By classifying the importance level of equipment operating status data and allocating energy level margin gradients according to equipment priority sequence to form differentiated energy level quotas, a hierarchical load chain is constructed to ensure differentiated energy security for equipment of different importance levels; combined with the hierarchical load chain to identify energy loss areas, and by analyzing the causes of loss to distinguish dispatchable losses, dispatch advantage areas are delineated based on recovery benefit values, and dispatch resources are concentrated in areas with higher benefits, improving the targeting of dispatch resource selection. 2. In terms of load balancing and resource management, load density analysis is used to form density fluctuation curves to identify transferable load sets, and a comprehensive response potential value is calculated by combining user response willingness and load transfer capacity to generate a load balancing flow; after extracting dispatch redundant resources from the load balancing flow, a resource association matrix is ​​constructed, shared loop resources are identified, and mutually exclusive resource pairs are calibrated by superimposed load calculation. After generating resource mutual exclusion relationship identifiers, adjustable load resource pools are formed according to response speed, and conflicting combinations are actively eliminated before dispatch execution, avoiding the risk of loop overload caused by parallel calls. 3. The scheduling execution adopts an architecture that combines regional autonomy with global coordination. The release priority is determined according to the energy early warning level. Target resource groups are extracted from the adjustable load resource pool and mutual exclusion checks are performed to form an effective resource set to build a dynamic scheduling network. The dynamic scheduling network is divided into multiple scheduling partitions. Each partition independently performs supply and demand balance analysis and constraint checks to generate partition feasible solutions. After being converted into partition scheduling configurations, global coordination and integration are performed to output scheduling control commands, enabling each region to have independent scheduling capabilities and improving the response speed and operational reliability of the scheduling system.

[0020] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0021] The accompanying drawings illustrate specific examples of the technical solutions described in this invention and, together with the detailed embodiments, form part of the specification, serving to explain the technical solutions, principles, and effects of this invention.

[0022] Unless otherwise specified, the same reference numerals in different figures represent the same or similar technical features, and different reference numerals may be used to represent the same or similar technical features.

[0023] Figure 1 This is a flowchart illustrating a method for intelligent monitoring and optimized scheduling of integrated energy resources in a park, as described in this invention.

[0024] Figure 2 This is a structural block diagram of a comprehensive energy intelligent monitoring and optimization scheduling system for industrial parks according to the present invention. Detailed Implementation

[0025] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, circuits, and methods are omitted so as not to obscure the description of this application with unnecessary detail.

[0026] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0027] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0028] The technical solutions of the embodiments of this application will be described below.

[0029] like Figure 1 As shown, this embodiment of the invention provides a method for intelligent monitoring and optimized scheduling of integrated energy in a park, including the following steps S110-S150:

[0030] Step S110: Collect comprehensive energy consumption data and equipment operation status data of the park, and construct a hierarchical load chain based on energy level matching of energy consumption data and equipment operation status data.

[0031] Specifically, comprehensive energy consumption data and equipment operation status data of the industrial park are collected. Smart metering terminals are deployed at each energy inlet and branch node of the park's comprehensive energy system. These terminals include electricity meters, gas meters, steam flow meters, and heating / cooling meters. The data collection frequency is set according to the energy type: electricity data is collected on a minute-by-minute basis, while gas and steam data are collected on an hourly basis. The smart metering terminals record the instantaneous power consumption, cumulative consumption, consumption time, and metering location of each energy medium, summarizing the data to form energy consumption data. Energy consumption data is categorized and labeled according to energy grade. For example, in a certain industrial park, an average daily electricity consumption of 12,000 kWh is labeled as high-grade energy, an average daily natural gas consumption of 800 cubic meters is labeled as high-grade energy, an average daily steam consumption of 15 tons is labeled as medium-grade energy, and an average daily chilled water supply of 200 tons is labeled as low-grade energy. Simultaneously, status monitoring devices are deployed on various energy-consuming equipment in the park to collect operating parameters such as equipment start / stop status, real-time load rate, cumulative runtime, and number of start / stop cycles. These devices upload the data to the energy management platform via an industrial bus or wireless network to form equipment operation status data. The equipment operation status data is indexed by the equipment number and associated with static attributes such as the region, function type, rated power and energy access type of the equipment. The equipment operation status data of this park includes 3 central air conditioning units, 50 data center server racks, 28 production line equipment and 12 lighting system zones.

[0032] In some embodiments, the step of constructing a hierarchical load chain by matching energy levels based on the energy consumption data and the equipment operating status data includes: classifying the equipment operating status data into equipment importance levels to obtain an equipment priority sequence; performing energy level margin gradient allocation based on the equipment priority sequence to obtain differentiated energy level quotas; performing supply and demand matching between the energy consumption data and the differentiated energy level quotas to form an energy level matching path; and constructing a hierarchical load chain based on the energy level matching path.

[0033] Equipment operation status data is used to classify equipment importance levels and obtain a priority sequence. Each equipment record in the operation status data is traversed to extract features such as functional attributes, service scope, and impact of interruptions. Based on functional attributes, equipment types are determined: core production equipment and security equipment are classified as Level 1 importance; office support equipment as Level 2 importance; and auxiliary service equipment as Level 3 importance. In the equipment operation status data, the data center server's functional attribute score F is 95 points, its service scope covering all 50 tenants in the park scores S 90 points, and the estimated daily loss of 80,000 yuan due to interruption results in a score I of 92 points. Based on the service scope, the area or number of people affected by equipment downtime is calculated; a larger affected area indicates higher importance. Based on the impact of interruptions, the economic losses and safety risks caused by equipment downtime are assessed; greater losses indicate higher importance. The importance score of each device in the equipment operation status data is calculated based on the above three dimensions. The scoring formula is P=W1×F+W2×S+W3×I, where P is the importance score, F, S, and I are the normalized scores of functional attributes, service scope, and interruption impact, respectively, with values ​​ranging from 0 to 100, and W1, W2, and W3 are weighting coefficients, with W1+W2W3=1. All devices are arranged in descending order of importance score to form an equipment priority sequence. In the equipment priority sequence, data center servers, fire pumps, and chiller units are at the front, while landscape lighting, non-essential air conditioning, and fountain pumps are at the back.

[0034] Differentiated energy level quotas are obtained through a gradient allocation of energy margins based on equipment priority sequences. The energy supply margin level available to each piece of equipment is determined by its ranking in the priority sequence. The top 20% of equipment in the priority sequence are allocated a Level 1 energy margin; for the 93 pieces of equipment in the park, the energy supply to the first 19 pieces of equipment is 120% of the rated demand, ensuring sufficient energy reserves for critical equipment during load fluctuations. Equipment in the 20% to 60% range of the priority sequence, i.e., equipment 20 to 56, are allocated a Level 2 energy margin, with an energy supply of 100% of the rated demand to maintain normal operation. The bottom 40% of the priority sequence, i.e., equipment 57 to 93, are allocated a Level 3 energy margin, with an energy supply of 80% to 100% of the rated demand, which can be moderately reduced during energy shortages. The energy level quota for each device is obtained by multiplying its rated energy consumption requirement by the corresponding margin factor. For example, a data center server with a rated power of 500kW multiplied by a margin factor of 1.2 yields a differentiated energy level quota of 600kW, while a landscape lighting device with a rated power of 30kW multiplied by a margin factor of 0.8 yields a differentiated energy level quota of 24kW. Since devices at different positions in the device priority sequence receive quotas with different margin factors, resulting in a gradient difference, this set of quotas constitutes the differentiated energy level quota. The differentiated energy level quota records the quota value and corresponding energy type of each device, indexed by its device number.

[0035] Energy consumption data is matched with differentiated energy level quotas to form an energy level matching path. Available energy and energy grade information for each energy supply node are extracted from the energy consumption data. The energy consumption data shows that the 10kV substation in the park can supply 8000kW of high-grade energy, the gas boiler can supply 5 tons / hour of steam (medium-grade energy), and the waste heat recovery device can supply 3 tons / hour of hot water (low-grade energy). The energy demand and energy level requirements of each device are extracted from the differentiated energy level quotas and matched sequentially from high to low energy level. First, high-grade energy supply nodes in the energy consumption data are paired with first-level margin equipment in the differentiated energy level quotas. The 10kV substation supplies 600kW of power to the data center server and 150kW to the fire pump, ensuring a high-quality energy supply for critical loads. Second, medium-grade energy supply nodes in the energy consumption data are paired with second-level margin equipment in the differentiated energy level quotas. The gas boiler supplies steam to the production line while simultaneously receiving waste energy from the high-grade energy utilization process. Finally, the low-grade energy supply nodes in the energy consumption data are paired with the three-level margin equipment in the differentiated energy level quota. Waste heat recovery hot water is used to supply heating to employee bathrooms and canteens, making full use of the waste energy at each level. During the matching process, the connection relationship, energy transfer amount, and energy level conversion between each pair of supply and demand nodes are recorded to form a supply and demand correspondence table. All the connection relationships in the supply and demand correspondence table are organized according to energy level hierarchy to form a directed path from high-energy level supply to low-energy level consumption. This set of paths is the energy level matching path, which describes the flow direction and transfer sequence from each energy node to each equipment node.

[0036] A hierarchical load chain is constructed based on energy level matching paths. The energy level hierarchy attributes of each path in the energy level matching path are analyzed, and paths at the same energy level are grouped into the same level. Paths directly supplying high-grade energy in the energy level matching path are assigned to the top layer, which includes power supply paths from the substation to the data center and from the substation to the production line, totaling 6500kW. Paths supplying medium-grade energy and primary cascade waste energy in the energy level matching path are assigned to the middle layer, which includes heating paths from boiler steam to process heating and cooling paths from air conditioning chilled water to the office area, totaling 2000kW equivalent. Paths supplying low-grade energy and secondary cascade waste energy in the energy level matching path are assigned to the bottom layer, which includes heating paths from waste heat hot water to domestic hot water and recovery paths from cooling tower heat dissipation to preheating, totaling 500kW equivalent. Within each level, paths are arranged from largest to smallest energy transfer capacity in the energy level matching path, with paths with large transfer capacity serving as the main paths of that level and paths with small transfer capacity serving as branch paths. Establish vertical connections between levels, connect the surplus energy output nodes of the upper path with the energy input nodes of the lower path to form a cross-level energy transfer channel, and integrate the three levels and their internal paths and inter-level connections into a unified hierarchical structure, which is the hierarchical load chain. The hierarchical load chain is presented in a tree structure with the root node as the total energy entrance of the park and each level of branches representing the energy distribution to different energy level loads.

[0037] Step S120: Identify energy loss areas by combining hierarchical load chains and equipment operating status data, and classify the energy loss areas according to the recoverability of losses to form scheduling advantage areas.

[0038] Specifically, energy loss zones are identified by combining hierarchical load chains and equipment operating status data. The energy transfer paths at each level of the hierarchical load chain are traversed, and the difference between the input and output energy for each path is calculated as the path loss, reflecting the degree of energy consumption during transmission. A correlation analysis is performed between the path loss in the hierarchical load chain and the operating efficiency of the corresponding equipment in the equipment operating status data. Parameters such as the actual operating efficiency, load rate, and runtime of the equipment are extracted from the equipment operating status data to assess the impact of equipment operating status on energy loss. In a certain industrial park, the air conditioning cooling path loss rate in the middle layer of the hierarchical load chain reached 20%, while the equipment operating status data showed that the industry benchmark loss rate for this type of chiller unit was 12%. The loss rate exceeding the standard by 8 percentage points indicates abnormal energy loss in this path segment. Further examination of the equipment operating status data revealed that the unit's long-term operation at a low load rate of 45% was the main cause of the high loss. When the loss rate of a certain path segment exceeds the industry benchmark, the area covered by that path segment is marked as a loss anomaly area. All loss anomaly areas in the hierarchical load chain are aggregated and spatially clustered using equipment location information from equipment operating status data. Adjacent loss anomaly areas are merged into continuous areas. Isolated loss points are grouped into separate areas based on the amount of loss. The continuous area formed after clustering is the energy loss zone. The energy loss zone records the area location, coverage area, total loss, and a list of associated equipment. In this park, three energy loss zones were identified, located in the refrigeration room area, along the steam pipeline, and in the lighting distribution area.

[0039] In some embodiments, the step of classifying the energy loss area based on its recoverability to form a scheduling advantage area includes: analyzing the causes of energy loss in the energy loss area to identify schedulable losses; evaluating the recovery costs of the schedulable losses to obtain recovery benefit values; prioritizing the recovery benefit values ​​to form a loss control sequence; and delineating the scheduling advantage area based on the loss control sequence.

[0040] Energy loss causes were analyzed to identify dispatchable losses in energy loss zones. For each energy loss zone, the specific causes of energy loss were analyzed, categorized into three types: equipment factors, operational factors, and environmental factors. Equipment factors include equipment aging, component wear, and design capacity mismatch; these losses require equipment replacement or modification and cannot be resolved through dispatching in the short term. Operational factors include frequent start-ups and shutdowns, low load rates, and deviations from optimal operating parameters; these losses can be reduced by adjusting operating strategies and are the primary targets for dispatching optimization. In a certain industrial park, the total energy loss in the chiller room's energy loss zone was 180kW. Cause analysis revealed: scale buildup on the chiller evaporators (50kW loss) was due to equipment factors, requiring shutdown for cleaning and not immediate dispatching; long-term low-load operation of a single chiller unit (85kW loss) was due to operational factors, which could be eliminated by switching to parallel operation of two units to increase the load rate; and damage to the chilled water pipe insulation (45kW loss) was due to environmental factors, requiring maintenance. Each loss in the energy loss zone is examined, and its cause is determined based on the equipment status, operating conditions, and environmental conditions. Losses caused by operational factors are extracted to form dispatchable losses. Dispatchable losses are recorded with loss location, loss amount, associated equipment, detailed cause, and suggested adjustment method. The dispatchable loss in this energy loss zone is 85kW, accounting for 47% of the total loss.

[0041] A recovery cost assessment is conducted to obtain a recovery benefit value for dispatchable losses. The assessment evaluates the costs and benefits of eliminating each dispatchable loss, calculating quantitative indicators of recovery benefits. Recovery costs include dispatch execution costs, user response compensation, equipment adjustment losses, and manual monitoring costs. Recovery benefits are the energy savings and reduced equipment wear after eliminating dispatchable losses. For example, the recovery operation for an 85kW dispatchable loss in a chiller room in an industrial park involves switching a single operating chiller unit to dual parallel operation. The operation requires maintenance personnel to monitor the startup status of the standby unit, adjust the chilled water flow valve opening to evenly distribute the load between the two units, and confirm stable operation before leaving the site. The entire switching process takes approximately one hour and involves manual monitoring and unit startup energy consumption. The recovery cost C and energy saving benefit B are estimated for each dispatchable loss. Both costs and benefits are converted to a uniform monetary unit (yuan) for comparison. The benefit value E = B / C is calculated. E is a dimensionless ratio; a larger ratio indicates better energy saving per unit of input and higher dispatch economics. The recovery benefit value records the benefit value of each loss and its calculation basis using the schedulable loss item as an index. A recovery benefit value greater than 1 indicates that the recovery benefit exceeds the recovery cost and is economically feasible, and should be included in the scheduling plan.

[0042] For example, the step of assessing the recovery cost of the schedulable loss to obtain a recovery benefit value includes: performing a recovery time window analysis on the schedulable loss to obtain an effective recovery period; estimating resource consumption during the effective recovery period to obtain a recovery cost; estimating energy savings after recovery of the schedulable loss to obtain energy-saving benefits; and calculating a recovery benefit value based on the energy-saving benefits, the recovery cost, and the effective recovery period.

[0043] Analysis of the recovery window for dispatchable losses is conducted to identify effective recovery periods. The time characteristics of each dispatchable loss are analyzed to determine the optimal execution time window for loss recovery operations. Some dispatchable losses only exist during specific time periods and require scheduling intervention during the loss occurrence to be effective. Recovery operations for some dispatchable losses are constrained by equipment operation and can only be performed during periods of low equipment load to avoid affecting normal production or service. For example, a certain industrial park has 85kW of dispatchable loss due to low-load operation of chiller units. Analyzing its time characteristics: In the morning, as cooling demand in the office area gradually increases, the unit load rate naturally rises to 70%, reducing the loss. However, it is not advisable to perform unit switching operations at this time to avoid affecting cooling stability. During the lunch break, cooling demand drops to a low point, and the unit load rate drops to 40%, making the loss apparent. Given the low cooling demand in the park, this is the ideal time to perform unit switching. Similarly, scheduling is not advisable after demand rebounds in the afternoon. By traversing the list of dispatchable losses and combining equipment operation plans, load forecast data, and process constraints, the time periods for executable recovery operations for each loss are identified. These identified time periods are divided into hours, and the recovery feasibility for each hour is marked to form effective recovery periods. Effective recovery periods are presented in the form of a timeline. Feasible periods are marked as effective, infeasible periods as invalid, and boundary periods as restricted. The longer the effective recovery period, the greater the flexibility of scheduling execution.

[0044] Resource consumption is estimated during the effective recovery period to obtain the recovery cost. The resources required to perform recovery operations for each schedulable loss within the effective recovery period are quantitatively estimated. The resources required for scheduling include labor monitoring costs, equipment adjustment energy consumption, and communication control costs; the resource composition varies depending on the type of recovery operation. The consumption of each resource is calculated based on the duration and type of the effective recovery period. For example, the effective recovery period for schedulable losses in a chiller room in an industrial park is during lunch break, which falls under off-peak electricity pricing. Switching the number of chiller units requires approximately one hour of on-site monitoring of the unit startup process, incurring labor costs. Starting the standby unit from a shutdown state to stable operation requires startup energy consumption, and issuing control commands and providing status feedback requires communication resources. The total cost required to perform the recovery of this loss is obtained by converting the resource consumption within the effective recovery period into a uniform currency unit and summing the results. The recovery cost is closely related to the selection of the effective recovery period. The same operation scheduled during peak electricity pricing increases equipment start-up and shutdown energy costs, thus increasing the recovery cost. Choosing off-peak electricity pricing within the effective recovery period can effectively reduce the recovery cost.

[0045] Energy savings are estimated after restoring dispatchable losses to obtain energy-saving benefits. The energy consumption reduction after eliminating various dispatchable losses is assessed and converted into economic benefits. The total energy savings are calculated based on the amount and duration of dispatchable losses. The energy saving formula is Q=P×T, where Q is the energy saving unit (kWh), P is the power loss of the dispatchable loss unit (kW), and T is the duration of the loss unit (h). The dimensional relationship is kW×h=kWh. In an industrial park, the dispatchable loss in a refrigeration room is 85kW. By switching from single-unit operation to dual-unit parallel operation, the load rate of each unit is increased from 45% to 70%, entering the high-efficiency range. The overall energy efficiency ratio of the units increases from 3.2 to 4.5. It is estimated that 60kW of the original 85kW loss can be eliminated. This energy-saving effect can continue until the end of the workday after the effective restoration period. The energy saving is calculated by multiplying the energy saved by the unit price of the corresponding energy source to obtain the economic value of energy saving, which is the energy saving benefit. The calculation of energy saving benefit takes into account the differences in energy types. Electricity is calculated based on time-of-use pricing, natural gas is calculated based on gas pricing, and steam is calculated based on heat pricing. There are differences in energy saving benefits between off-peak and peak electricity pricing periods. Energy saving benefit reflects the potential economic value of recovery operations.

[0046] The recovery benefit value is calculated based on the combination of energy-saving benefits and recovery costs with the effective recovery period. The energy-saving benefits are divided by the recovery costs to obtain the basic benefit ratio, which is then adjusted according to the characteristics of the effective recovery period to obtain a comprehensive benefit index reflecting economic efficiency and feasibility. The basic benefit ratio formula is R=B / C, where R is dimensionless, B is the energy-saving benefit unit, and C is the recovery cost unit. When the effective recovery period is short and the time constraint for scheduling is tight, a timeliness correction coefficient k=T_eff / T_ref is introduced, where T_eff is the effective recovery period length in hours (h), T_ref is a reference duration of 8 hours, and k is a dimensionless coefficient. The recovery benefit value is calculated as E=R×(1+k)=(B / C)×(1+T_eff / T_ref), introducing the (1+k) form to prevent excessive penalty for the benefit value in short windows. In a certain industrial park, the centrifugal chiller units in the refrigeration room operate in single-unit mode during the summer. Due to the sudden drop in cooling load caused by employees leaving for lunch in the office buildings at midday, each unit operates at low load for an extended period, resulting in dispatchable losses. The recovery operation for this loss involves switching from single-unit to parallel operation during the lunch break, restoring the load rate of each unit to its high-efficiency range. Since the lunch break coincides with off-peak electricity pricing and low cooling demand, the recovery cost is low, and the energy-saving benefits can be extended to the afternoon peak period. Considering all factors, the recovery benefit of this dispatchable loss ranks first in the park and should be prioritized in the dispatching plan. Dispatchable losses caused by improper valve opening along the steam pipeline have higher recovery costs due to the complexity of coordinating adjustments to multiple regulating valves involved in the recovery operation. Dispatchable losses caused by imprecise zoning control in the lighting distribution area require modification of the time control strategy, resulting in a longer implementation cycle and limited effective recovery time. The recovery benefits of these two types of losses decrease sequentially.

[0047] A loss control sequence is formed by prioritizing losses based on their recovery benefit values. All dispatchable losses are sorted from highest to lowest according to their recovery benefit value, with the highest recovery benefit value ranked first, followed by the next highest. In a certain industrial park, three energy loss zones were identified as dispatchable losses. The low-load loss of the chiller units in the refrigeration room has the highest recovery benefit value and is prioritized first. This loss can be eliminated through optimized scheduling of the number of units, with significant energy-saving effects and low implementation difficulty. The loss due to improper valve opening in the steam pipeline has the second highest recovery benefit value and is ranked second. This loss requires adjusting the opening of multiple regulating valves along the pipeline to match the steam flow with the end-point demand, involving multi-point coordination. The loss due to imprecise zoning control in the lighting distribution area has the lowest recovery benefit value and is ranked third. This loss requires refined lighting zoning and adjustment of the time control strategy, with a longer implementation cycle. During the sorting process, losses with similar recovery benefit values ​​are further sorted, with the magnitude of the loss as a secondary sorting criterion. Losses with larger magnitudes are prioritized to ensure that high-impact losses are addressed first. The sorting results form an ordered list. Each entry in the list contains a loss number, associated equipment, loss amount, recovery benefit value, suggested recovery period, and expected energy saving. This ordered list is the loss control sequence. The loss control sequence, from beginning to end, represents the priority order of scheduling optimization and is the core basis for scheduling decisions.

[0048] Dispatch advantage areas are delineated based on loss control sequences. The top-ranked loss items are selected from the loss control sequence, and their corresponding physical areas are marked on the park's floor plan based on the loss location information recorded in each item. The selection ratio is set to the top 50% of the loss control sequence. For example, in an industrial park with three loss control records, the top two are selected: chiller room loss and steam pipeline loss. The selection ratio can be adjusted according to the availability of dispatch resources; when resources are plentiful, the ratio can be increased to expand the coverage area, while when resources are scarce, the ratio can be decreased to focus on core areas. The physical areas corresponding to the selected items in the loss control sequence are marked on the park's floor plan. The chiller room area, containing three chiller units and six chilled water pumps, is the core hub of the park's cooling energy. The steam pipeline area, containing twelve regulating valves and eight steam traps, is a key channel for heat energy transmission and distribution in the park. These two areas are not spatially adjacent and form independent advantage area boundaries. The area within the boundary is the scheduling advantage area. The scheduling advantage area is the spatial range with the highest scheduling efficiency density in the park, which concentrates multiple high-efficiency scheduling losses. The scheduling advantage area is stored in the form of geographic information, including the area boundary coordinates, coverage area, number of loss entries, and expected total energy saving.

[0049] Step S130: Perform load density analysis on the scheduling advantage area to form a density fluctuation curve, identify the transferable load set based on the density fluctuation curve, and perform user-side response potential assessment based on the transferable load set to generate a load balancing flow.

[0050] Specifically, load density analysis is performed on the dispatch advantage area to form a density fluctuation curve. The total energy load at each moment within the dispatch advantage area is statistically analyzed, and real-time power data collected from various metering points within the area is aggregated to obtain regional-level load time-series data. Dividing the total load by the area of ​​the dispatch advantage area yields the load density per unit area. This load density eliminates the influence of regional area differences, facilitating horizontal comparisons between different areas. In a certain industrial park, the dispatch advantage area of ​​the refrigeration room experiences a rapid increase in load density during the morning rush hour due to the concentrated arrival of office workers and the use of air conditioning and office equipment. At noon, a large number of people leave for lunch, causing the air conditioning load to automatically decrease, bringing the load density back down to half of the morning peak. In the afternoon, after people return, the load density rises again, forming a bimodal characteristic. Connecting the load density values ​​at each moment in chronological order forms a density fluctuation curve. The density fluctuation curve, with time on the horizontal axis and load density on the vertical axis, reflects the temporal distribution characteristics of the load within the dispatch advantage area. The peak value of the density fluctuation curve corresponds to the peak energy consumption period, and the trough value corresponds to the low energy consumption period. Analyzing the fluctuation amplitude and frequency of the density fluctuation curve reveals that a larger fluctuation amplitude indicates greater peak-shaving and valley-filling potential within the dispatch advantage area, making it more suitable for load transfer dispatching.

[0051] Identifying Transferable Load Sets Based on Density Fluctuation Curves. The morphological characteristics of the density fluctuation curves are analyzed to identify peak and trough periods. Peak periods are those where the load density exceeds the daily average by a certain percentage. Various loads operating during the peak periods of the density fluctuation curves are statistically analyzed, and load types with time flexibility are selected. Time flexibility refers to the ability of a load's operating time to be adjusted within a certain range without affecting its functionality. For example, the density fluctuation curve of an industrial park shows the afternoon as the peak period. Loads operating during this period include central air conditioning, office lighting, data centers, and electric vehicle charging stations. Central air conditioning, due to thermal inertia, can pre-cool during trough periods, storing cooling capacity in the building envelope to reduce cooling output during peak periods. Charging stations serve commuter vehicles with ample parking time at night, allowing charging to be delayed until the lower-price evening hours. Data centers, carrying the park's information services, are continuously operating; interruptions to this load would disrupt tenants' normal operations and are therefore not transferable. Loads assessed as transferable are aggregated to form a transferable load set. This set records the load number, type, rated power, transferable period, and transfer constraints for each load, representing a subset of the loads from the peak periods of the density fluctuation curve that meet the transfer conditions.

[0052] In some embodiments, the step of generating a load balancing flow by assessing user-side response potential based on the transferable load set includes: collecting user response intention data from the transferable load set to obtain a response intention level; performing transfer capacity statistics on the transferable load set to obtain a load transfer capacity; calculating a comprehensive response potential value based on the response intention level and the load transfer capacity; and generating a load balancing flow based on the comprehensive response potential value.

[0053] User willingness to participate in load transfer is collected to obtain a willingness rating for the transferable load set. For each user belonging to a load in the transferable load set, their willingness to participate in load transfer is collected to assess their level of cooperation with dispatching. Willingness collection methods include historical response record analysis and real-time willingness query. Historical response record analysis extracts user response rates and response stability from past dispatching events, while real-time willingness query collects user participation intentions for the current dispatch through user terminals. In an industrial park, the air conditioning load of Building A office building in the transferable load set is managed by the property management company. The property management company has signed a demand response agreement with the park and has assigned a dedicated energy manager. In the past six months, the company has responded actively and implemented the dispatching requests effectively. Based on this, the willingness rating for this load is 0.85. The users of the charging pile load in Building B are the private car owners of tenant employees. Due to the uncertainty of car owners' usage times caused by personal matters and the lack of a unified coordination mechanism, historical dispatching often shows situations where car owners need to use their vehicles temporarily, interrupting the response. The willingness rating for this load is 0.35. The response willingness score ranges from 0 to 1. Each load in the transferable load set inherits the response willingness score of its user as a reference for subsequent scheduling and allocation.

[0054] The load transfer capacity is obtained by statistically analyzing the transfer capacity of the transferable load set. Based on each load in the transferable load set, the maximum transfer capacity of each load is calculated to assess its upper limit of regulation potential. The transfer capacity is limited by factors such as the regulation range of the load equipment, the user's allowed regulation range, and safe operation constraints. In an industrial park, the air conditioning load of Building A in the transferable load set uses an ice storage air conditioning system. During off-peak hours at night, the ice-making mode can be turned on to store cold energy in the ice storage tank. During peak hours during the day, the ice is melted to release cold energy and reduce the operation of the main unit. The capacity of the ice storage tank determines the upper limit of the load transfer capacity of this load. The charging pile load of Building B serves employee commuting vehicles. Vehicles usually connect to the charging pile after get off work in the afternoon and leave the next morning. Under the premise of meeting the travel needs of the next morning, the charging time at night can be flexibly adjusted. The duration of delayed charging determines the load transfer capacity of this load. The transfer capacity of each load in the transferable load set is summarized to form a capacity list indexed by the load number, which is the load transfer capacity. The unit of load transfer capacity is kWh, representing the regulation power that each load can contribute.

[0055] The comprehensive response potential value is calculated based on response willingness and load transfer capacity. Multiplying the response willingness by the load transfer capacity yields the effective response capacity for each load, i.e., the actual capacity that users are willing and able to participate in the transfer. The formula for effective response capacity is Q_eff = W × Q, where Q_eff is the effective response capacity in kWh, W is the dimensionless response willingness, and Q is the load transfer capacity in kWh. A response reliability coefficient r is introduced to correct for the effective response capacity. The response reliability coefficient is determined based on the historical completion rate of user responses. The comprehensive response potential value is calculated as V = W × Q × r. In an industrial park, the air conditioning load of Building A is managed by a professional maintenance team. Historically, its scheduling execution completion rate is high, and there have been no instances of mid-process withdrawal. Its response willingness is high, its load transfer capacity is large, and its reliability coefficient is close to full value. All three indicators are at an excellent level, placing its comprehensive response potential value among the top in the park. While the load transfer capacity of the charging piles in Building B is comparable to that of Building A's air conditioning, lower response willingness due to individual differences among car owners and temporary interruptions in historical execution lower the reliability coefficient, resulting in a significantly lower comprehensive response potential value than Building A's air conditioning. The comprehensive response potential value comprehensively reflects the response capabilities of various loads in terms of willingness, capacity, and reliability, and is a core indicator for evaluating the value of load scheduling.

[0056] A load balancing flow is generated based on the comprehensive response potential value. Loads in the transferable load set are sorted from highest to lowest comprehensive response potential value, with high-potential loads prioritized for load balancing. The total target load to be transferred is determined based on the peak-to-valley difference of the density fluctuation curve. Loads are selected sequentially from the sorted load list, and their comprehensive response potential values ​​are accumulated until the target value is reached. In an industrial park, after sorting by comprehensive response potential value, air conditioning loads in buildings under unified property management are ranked higher due to their mature response mechanisms and reliable execution, while charging piles and tenant-owned equipment under decentralized management are ranked lower due to coordination difficulties. Prioritizing the higher-ranked air conditioning loads for this load balancing dispatch can meet the peak shaving target. Charging pile loads are not included in this dispatch but are retained in the reserve list for future expansion. Specific transfer time periods and transfer volumes are assigned to each selected load, and the transfer arrangements of all selected loads are integrated to form a load balancing flow. The load balancing flow is presented as a flow diagram describing the migration path and amount of each load from its original operating period to its target operating period. After the implementation of the load balancing flow, the peak-to-valley difference of the density fluctuation curve can be reduced, improving the stability of the park's load.

[0057] Step S140: Perform scheduling capacity analysis on the load balancing flow to extract scheduling redundant resources, perform call conflict analysis on the scheduling redundant resources to obtain resource mutual exclusion relationship identifiers, and form an adjustable load resource pool based on the resource mutual exclusion relationship identifiers and hierarchical response speed.

[0058] Specifically, a scheduling capacity analysis is performed on the load balancing flow to extract scheduling redundancy resources. The transfer arrangements of each load in the load balancing flow are analyzed, and the capacity difference between the total capacity of the loads involved in the transfer and the actual capacity called up in each time period is calculated. Load transfer schemes in the load balancing flow typically reserve a certain capacity margin to cope with execution deviations. The remaining adjustable capacity is obtained by subtracting the actual allocated transfer amount from the comprehensive response potential value of each load in the load balancing flow. In a certain industrial park's load balancing flow, the ice storage tank capacity of the air conditioning load in Building A is sufficient. However, this scheduling only requires reducing a portion of the peak load and does not require using all the ice storage capacity. Therefore, the actual allocated transfer amount is less than its comprehensive response potential value, and the remaining ice storage capacity is the remaining adjustable capacity of this load, which can be further called up in case of emergencies. A scheduling redundancy resource list is formed by summarizing all loads with remaining adjustable capacity in the load balancing flow. The scheduling redundancy resources record the load number, type, remaining capacity, region, and available time period for each load. This is the backup adjustment capacity that can still be used after the load balancing flow is executed, and it can be quickly called up in case of sudden supply-demand gaps or early warning events.

[0059] In some embodiments, the step of performing call conflict analysis on the scheduled redundant resources to obtain resource mutual exclusion relationship identifiers includes: performing device correlation analysis on the scheduled redundant resources to obtain a resource correlation matrix; identifying shared loop resources and independent loop resources based on the resource correlation matrix; performing parallel call constraint detection on the shared loop resources to obtain mutually exclusive resource pairs; and generating resource mutual exclusion relationship identifiers based on the mutually exclusive resource pairs and the independent loop resources.

[0060] A resource correlation matrix is ​​obtained by performing equipment correlation analysis on redundant scheduling resources. Based on each resource in the redundant scheduling resource list, the electrical wiring diagram, control system topology, and physical location information of the equipment to which each resource belongs are extracted. The correlation between any two redundant scheduling resources is analyzed to determine if there is a correlation. Correlation relationships include shared power supply circuits, shared controllers, and physical spatial adjacency. Resource calls within shared power supply circuits can affect each other's circuit loads, and resource control commands within shared controllers may cause timing conflicts. In a certain industrial park office building, the air conditioning and lighting on the third floor both draw power from the same distribution cabinet in the floor's distribution room. When both loads increase their output simultaneously, the current superimposed flowing through the same incoming switch poses an overload risk, indicating a shared power supply circuit correlation. However, the charging piles in the underground parking garage are powered by independent dedicated charging transformers and are electrically isolated from the office building loads, so there is no correlation. The correlation analysis results are constructed into a two-dimensional matrix, where the rows and columns correspond to the numbers of the redundant scheduling resources. The matrix element values ​​represent the correlation strength and type of the corresponding resource pairs, forming the resource correlation matrix. Resource pairs with high correlation strength in the resource correlation matrix require close attention to the feasibility of their parallel calls.

[0061] Resource association matrices are used to identify shared and independent circuit resources. Each row of the resource association matrix is ​​scanned to count the number and type of associations between each resource and other resources. If all association elements of a resource in the resource association matrix are zero, it indicates that the power supply circuit and control channel of that resource are independent of other resources, and it is classified as an independent circuit resource. If a resource has non-zero association elements in the resource association matrix, it indicates that the resource shares a certain circuit with other resources, and it is classified as a shared circuit resource. Analysis of the resource association matrix of an industrial park shows that the charging piles in the underground parking garage are powered by dedicated transformers and equipped with independent charging management controllers. They have no electrical or control associations with other loads in the park and are classified as independent circuit resources. When scheduling these resources, coordination with other resources is not required, and they can be called in parallel with any other resource. Air conditioning, lighting, sockets, and other loads on each floor of the office building share the floor distribution box and building automation system, and are classified as shared circuit resources. When scheduling these resources, it is necessary to assess whether parallel calls exceed the carrying capacity of the shared circuit. Shared circuit resources require further testing for parallel call constraints, while independent circuit resources have a higher degree of scheduling freedom and can be directly included in the callable scope.

[0062] For example, the step of performing parallel call constraint detection on the shared loop resources to obtain mutually exclusive resource pairs includes: extracting the loop capacity of the shared loop resources to obtain the loop carrying capacity limit; performing call power statistics on the shared loop resources to obtain the resource call power; performing superimposed calculation based on the loop carrying capacity limit and the resource call power to obtain the superimposed load value; and performing over-limit judgment and calibrating the mutually exclusive resource pairs based on the superimposed load value.

[0063] The circuit capacity is extracted from shared circuit resources to obtain the upper limit of circuit load. The rated capacity parameters of the circuits connected to the shared circuit resources are extracted, and the minimum value of each constraint, considering factors such as cable current carrying capacity, switch rating, and transformer capacity, is taken as the effective capacity of the circuit. For each power supply circuit involving shared circuit resources, its design capacity and actual operating capacity are reviewed. Cable current carrying capacity is determined based on cable cross-section and laying method; switch rating is determined based on circuit breaker nameplate parameters; and transformer capacity is determined based on the capacity share allocated to the circuit in the power distribution design. For example, in the floor distribution boxes connected to the shared circuit resources on the third floor of an office building in an industrial park, the power distribution system archives show that the transformer allocated capacity is the smallest among the incoming cable current carrying capacity, incoming switch rating, and upstream transformer allocated capacity. Therefore, the transformer allocated capacity of 100kW is taken as the effective capacity. Multiplying the effective capacity by a safety factor yields an upper limit of circuit load of 85kW. The upper limit of circuit load is the benchmark value for judging the feasibility of parallel use of shared circuit resources. Exceeding the upper limit of circuit load will cause the upstream transformer to overheat or trigger the incoming switch protection, resulting in circuit tripping. A mapping table is established between shared circuit resources and their corresponding circuit capacity limits. Multiple shared circuit resources under the same circuit share the same circuit capacity limit value. The mapping table supports quick querying of the capacity constraints of any shared circuit resource.

[0064] The power consumption of shared circuit resources is statistically analyzed to obtain the resource call power. The system iterates through each resource in the shared circuit resource list, calculating the power requirement of each resource during scheduling. Resource call power refers to the actual electrical power consumed when a resource participates in a scheduling response. The rated power and adjustable range of each resource are extracted from the attribute data of the shared circuit resources to calculate the resource call power at the maximum call depth. For air conditioning loads, the call power is the full-load power under cooling or heating conditions; for lighting loads, the call power is the total power when all lights are on; and for socket loads, the call power is statistically calculated based on historical peak power consumption. The shared circuit resources on the third floor of an office building in an industrial park include three categories: air conditioning terminals, lighting fixtures, and socket circuits. The air conditioning terminals, including multiple fan coil units and fresh air units, have the highest resource call power when running at full load under summer cooling conditions. The lighting fixtures cover open office areas, meeting rooms, and corridors; when all are on, the resource call power depends on the number of fixtures. The socket circuits supply computers and office equipment at each workstation; the resource call power during peak hours is calculated based on historical peak power. Establish a correspondence table between shared circuit resource numbers and resource access power. Resource access power is the basis for subsequent calculation of superimposed load. The higher the power of the shared circuit resource, the more it occupies the circuit's upper limit.

[0065] The superimposed load value is obtained by superimposing the circuit capacity limit and the resource call power. For shared circuit resource combinations, the circuit capacity limit is used as the capacity benchmark for calculation, and the resource call power of each resource in the combination is added to calculate the total power demand during parallel call. The superimposed calculation uses direct power addition, which is suitable for purely resistive loads. For inductive loads containing motors, the effects of power factor and starting current must also be considered, and apparent power should be used for superposition. The maximum load capacity of a shared circuit on the third floor of an office building in an industrial park is 85kW. During the hot summer afternoons, the air conditioning units need to operate at full capacity, requiring 45kW of power. On cloudy days with insufficient sunlight, all lighting fixtures need to be turned on, requiring 25kW of power. When these two shared circuit resources operate simultaneously, their respective power consumption is added together, resulting in a superimposed load of 70kW, which is 82% of the circuit's maximum load capacity and within the safe operating range. However, if this coincides with peak employee attendance and all workstation computers are running simultaneously, the power consumption increases to 30kW. The superimposed load of all three shared circuit resources operating simultaneously rises to 100kW, exceeding the circuit's maximum load capacity by 118%. This exceeds the capacity constraint. The solution involves iterating through all pairwise and multi-resource combinations within the same shared circuit, calculating the superimposed load value, and comparing it to the circuit's maximum load capacity to obtain the load percentage. The results are then recorded in a table showing each combination, its superimposed load value, and its load percentage. The superimposed load value reflects the total load pressure on the loop when resources are called in parallel. The load ratio intuitively shows the margin level of the superimposed load value relative to the loop's capacity limit. The closer the load ratio is to or exceeds 100%, the higher the risk of parallel calls.

[0066] The system uses superimposed load values ​​to determine whether mutually exclusive resource pairs are exceeded. The superimposed load value of each resource combination is compared with the corresponding circuit's upper limit to determine if it exceeds the limit. If the superimposed load value is less than or equal to the circuit's upper limit, it indicates that the parallel use of this resource combination is not mutually exclusive within the circuit's capacity and can be used simultaneously. If the superimposed load value is greater than the circuit's upper limit, it indicates that the parallel use of this resource combination will exceed the circuit's capacity and constitutes a mutually exclusive relationship, requiring it to be designated as a mutually exclusive resource pair. In an industrial park office building, the upper limit of the circuit's capacity on the third floor is 85kW. The superimposed load value of the shared circuit resources for air conditioning and lighting is less than the circuit's upper limit, indicating they are not mutually exclusive and can be used simultaneously. In actual operation, during summer cloudy days, both often operate in parallel without tripping, verifying this determination. However, when air conditioning, lighting, and sockets are all running at full load simultaneously, the superimposed load value exceeds the circuit's upper limit. During a departmental training session, the main switch of the floor's distribution box tripped due to all meeting room lights being on, air conditioning being at full cooling power, and a large number of laptops being plugged into sockets simultaneously. Subsequent analysis confirmed that all three being on constitutes a mutually exclusive resource pair. Resource combinations that are deemed to have exceeded the limit are marked as mutually exclusive resource pairs. The extent of the limit exceedance, typical scenarios that trigger the limit exceedance, and suggested staggered peak call measures are recorded for each mutually exclusive resource pair.

[0067] Resource mutual exclusion relationship identifiers are generated based on mutually exclusive resource pairs and independent loop resources. The information of mutually exclusive resource pairs is organized into a standardized mutual exclusion identifier format, which includes five fields: mutual exclusion identifier code, mutual exclusion type, involved resource number, over-limit range, and recommended measures. The mutual exclusion identifier code adopts the encoding rule "MX-loop number-serial number". The mutual exclusion type is divided into two categories: capacity mutual exclusion and timing mutual exclusion. Capacity mutual exclusion refers to parallel calls exceeding the loop's capacity limit, while timing mutual exclusion refers to timing conflicts in control commands. For example, a combination of air conditioning, lighting, and sockets all on on the third floor of an office building in an industrial park is identified as a mutually exclusive resource pair, assigned the mutual exclusion identifier MX-3F-001, with a mutual exclusion type of capacity mutual exclusion. The involved resource numbers are AC-3F, LT-3F, and SK-3F. The over-limit range is 15kW, which is the difference between the superimposed load value of 100kW and the loop's capacity limit of 85kW. The recommended measure is to stagger the use of air conditioning and sockets at intervals of no less than 30 minutes. Independent loop resources are marked as non-mutually exclusive, their mutual exclusion identifiers are set to null, and their mutual exclusion type is marked as "independent," indicating that the resource can be used in parallel with any other resource without conflict checking. Underground parking garage charging piles, as independent loop resources, have null mutual exclusion identifiers, independent mutual exclusion types, and an empty list of mutual exclusion objects, giving them the highest degree of freedom in scheduling. A complete resource mutual exclusion relationship identifier dataset is constructed by aggregating the identifiers of all mutual exclusion resource pairs and independent loop resources. This dataset records the mutual exclusion identifier, mutual exclusion object list, and mutual exclusion details for each resource, indexed by its resource number. It supports quick querying of mutual exclusion constraints for any resource and is the core basis for scheduling resource selection and conflict avoidance.

[0068] Adjustable load resource pools are formed based on resource mutual exclusion identifiers and tiered by response speed. Redundant resources are grouped according to these identifiers, and mutually exclusive resources are allocated to different call batches to ensure parallel access within the same batch. Based on these groupings, resources are tiered according to their response speed, which refers to the time required for a resource to complete load adjustment from receiving a scheduling command. For example, an industrial park's energy storage battery uses a power electronic converter for control, requiring only a few seconds from receiving a scheduling command to completing power adjustment without manual intervention; this is categorized as a fast response level suitable for emergency scheduling scenarios such as grid frequency fluctuations. Air conditioning pre-cooling requires starting the refrigeration unit in advance and waiting for the ice storage tank or building's cooling to reach the expected level; it takes tens of minutes from the issuance of the command to the formation of effective regulation capacity; this is categorized as a normal response level suitable for day-ahead or intraday scheduling scenarios. Delayed charging at charging piles involves coordinating vehicle owners' travel plans, requiring advance notification and owner confirmation; this is categorized as a delayed response level suitable for pre-planned scheduling scenarios. The tiered resources are organized into a hierarchical adjustable load resource pool. The adjustable load resource pool contains resource subsets with three response levels. The resources in each subset have excluded conflicting combinations in the resource mutual exclusion relationship identifier and can be directly called in parallel. It is a structured collection of all adjustable resources available in the park.

[0069] Step S150: For the load balancing flow, perform supply and demand risk detection to obtain the energy early warning level, and trigger multi-level linkage to generate a dynamic scheduling network for the adjustable load resource pool according to the energy early warning level.

[0070] Specifically, supply and demand risk detection is conducted to obtain energy early warning levels for load balancing flows. The execution of load balancing flows is monitored, comparing actual load transfer with planned transfer to assess the deviations in the execution of various transfer schemes within the load balancing flow. When the actual transfer is lower than the planned value, it indicates that some load has not been transferred according to the load balancing flow plan, the peak shaving effect has not met expectations, and may lead to supply and demand gaps during peak periods. During the summer peak electricity consumption period, the air conditioning pre-cooling transfer scheme arranged in the load balancing flow only completed 60% of the planned transfer due to a temporary failure of the refrigeration unit. This, coupled with a temporary power rationing notice from the power grid requiring a 15% load reduction, resulted in a sharp drop in available capacity on the supply side. In this case, it is necessary to comprehensively assess the combined impact of load balancing flow execution deviations and supply-side risk factors. Simultaneously, the real-time status of the park's energy supply side is monitored, including grid power supply capacity, distributed power output, and energy storage charge status, to identify potential risks on the supply side. The load balancing flow execution deviations and supply-side risk factors are comprehensively analyzed to calculate the supply-demand balance margin, which is the difference between supply capacity and predicted load. Energy warning levels are determined based on the magnitude of the supply-demand balance margin. A margin greater than 20% indicates no warning; a margin between 10% and 20% indicates a Level 3 warning; a margin between 5% and 10% indicates a Level 2 warning; and a margin less than 5% indicates a Level 1 warning. The energy warning level reflects the degree of tension in the park's energy supply and demand. Higher levels indicate a more pronounced supply-demand imbalance, necessitating the activation of emergency dispatch measures. Energy warning levels are updated in real time; when load balancing improves or supply-side risks are eliminated, the energy warning level is lowered accordingly.

[0071] In some embodiments, the step of triggering the adjustable load resource pool to generate a dynamic scheduling network based on the energy warning level through multi-level linkage includes: determining a release priority based on the energy warning level; extracting a target resource group from the adjustable load resource pool according to the release priority; performing mutual exclusion checks on the target resource group to eliminate conflicting resources and form an effective resource set; and constructing a dynamic scheduling network based on the effective resource set.

[0072] Release priorities are determined based on energy warning levels. A mapping rule is established between energy warning levels and resource release strategies, with different energy warning levels corresponding to different release priorities and ranges. Level 1 warnings correspond to the highest release priority, prioritizing the release of fast-response resources, followed by conventional-response resources, and finally delayed-response resources, to quickly fill the supply-demand gap. For example, if an industrial park experiences temporary power rationing in the afternoon during summer, causing the supply-demand balance margin to plummet from 12% to 3%, the energy warning level jumps from Level 3 to Level 1. According to the release priority determined by the mapping rule, fast-response resources such as energy storage batteries are deployed within 30 seconds to provide power support, while conventional-response resources such as air conditioning load reduction are included in the next batch of release queues. Level 2 warnings correspond to a medium release priority, prioritizing the release of conventional-response resources, with fast-response resources as reserves and delayed-response resources in standby mode. Level 3 warnings correspond to a lower release priority, releasing only delayed-response resources for preventative scheduling, while fast-response and conventional-response resources remain on standby. In the absence of a warning, the release priority is empty, and no resource release is triggered. Release priority is represented by a priority code, which includes parameters such as release order, release ratio, and time window. Release priority ensures appropriate response levels at different urgency levels, avoiding resource waste from over-response or supply-demand imbalance from under-response. When the energy alert level is upgraded from Level 3 to Level 2, the release priority is increased accordingly, and previously standby conventional response-level resources are placed in the release queue.

[0073] Target resource groups are extracted from the adjustable load resource pool according to release priority. Resources are selected sequentially from each response level of the adjustable load resource pool according to the release order of release priority. First, the capacity requirement of the current supply-demand gap is determined as the target capacity for resource selection. Resources are selected from the resource subset of the corresponding response level of the adjustable load resource pool in the order specified by the release priority, prioritizing resources with large remaining adjustable capacity and fast response speed. For example, an industrial park has a 150kW supply-demand gap under a Level 2 warning. Resources are first screened from the regular response level of the adjustable load resource pool according to release priority. Building A, with 80kW of remaining air conditioning pre-cooling capacity and high property management cooperation, is selected first. Building B, with 50kW of charging delay capacity but historically low owner response rate, is selected second. When the cumulative capacity of the regular response level resources is still insufficient, 30kW of non-essential lighting dimming is selected from the delayed response level of the adjustable load resource pool as a supplement. The capacity of the selected resources is accumulated during the selection process, and selection stops when the cumulative capacity reaches or exceeds the target capacity. The selected resources are summarized to form a target resource group, which is the set of resources to be called up under the current warning level. The total capacity of the target resource group should be slightly larger than the supply-demand gap, with a 10%-15% margin reserved to account for execution deviations. The target resource group records the resource number, capacity, response level, and estimated call time for each resource. When extracting resources from the adjustable load resource pool, a dual principle of capacity priority and response speed priority should be followed to ensure that the target resource group has sufficient adjustment capacity and response speed.

[0074] The target resource group undergoes mutual exclusion checks to eliminate conflicting resources and form a valid resource set. The resource mutual exclusion identifier is invoked to check for mutual exclusion conflicts among resources in the target resource group. All resources in the target resource group are paired, and the corresponding mutual exclusion record is checked against the resource mutual exclusion identifier. For example, in an industrial park, the target resource group includes a 45kW air conditioning load and a 25kW socket load on the 3rd floor of an office building. The mutual exclusion identifier reveals that both share the same floor's power distribution circuit with a circuit capacity limit of only 85kW. A tripping occurred due to simultaneous full-load operation, indicating a mutual exclusion record, and one of them needs to be excluded. If a resource pair is mutually exclusive, one resource needs to be excluded from the target resource group. The exclusion principle is to retain resources with larger capacity or faster response speed. The 3rd floor air conditioning load, with its large capacity and mature response mechanism, is retained, while the 3rd floor socket load is excluded. For cases where multiple resources exhibit chained mutual exclusion, a greedy algorithm is used to select the combination with the largest total retained resource capacity. The remaining resource set after excluding conflicting resources forms a valid resource set. Resources in the valid resource set do not have mutual exclusion relationships and can be safely used in parallel. The total capacity of the effective resource set may be less than the target resource set after excluding conflicting resources. If the capacity is insufficient to fill the supply-demand gap, non-mutually exclusive resources need to be added from the adjustable load resource pool. Lighting or air conditioning loads that do not have a mutual exclusion relationship with existing resources can be selected from other floors to supplement the resources. The effective resource set records the number, confirmed capacity, and call parameters of each resource and is the final list of resources participating in scheduling.

[0075] A dynamic scheduling network is constructed based on an effective resource set. Resources in the effective resource set are used as network nodes, and connections between nodes are established according to the supply and demand matching relationship of resources. Supply-side resource nodes are connected to demand-side load nodes, and the connection edges represent the direction and magnitude of energy or power flow. The geographical location and electrical access points of each resource in the effective resource set are analyzed, and resources that are geographically close or on the same circuit are organized into local subnetworks. For example, in an industrial park, the air conditioning and lighting resources of Building A in the effective resource set are located in the same building and connected to the same power distribution room, forming a subnetwork for Building A with internal transmission loss of less than 1%. The charging piles and energy storage resources in the underground parking garage share a dedicated charging transformer, forming a charging subnetwork. All subnetworks are interconnected through the park's 10kV main power distribution network. When the Building A subnetwork lacks its own resources, it can obtain energy storage support from the charging subnetwork through the main network. The local subnetworks are connected through the main network to form a hierarchical dynamic scheduling network topology. The node attributes of the dynamic scheduling network include resource number, resource type, adjustable capacity, and response status; the edge attributes include flow magnitude, transmission loss, and priority. The dynamic scheduling network supports network flow algorithms for optimized scheduling, calculating the optimal call volume for each node and the optimal traffic allocation for each edge. Stored in a graph data structure, the dynamic scheduling network supports dynamic addition and deletion of nodes and edges, adapting to network reconstruction needs when warning levels change. When visualized, nodes are represented by different colors to indicate resource types, and edge thickness represents traffic volume, facilitating intuitive understanding of the scheduling scheme by operations personnel.

[0076] Step S160: Based on the dynamic scheduling network, perform partitioned scheduling to generate partitioned scheduling configurations, and perform global coordination and integration of the partitioned scheduling configurations to output scheduling control commands.

[0077] In some embodiments, generating a partitioned scheduling configuration based on the dynamic scheduling network includes: dividing the dynamic scheduling network into regional boundaries to obtain scheduling partitions; performing supply and demand balance analysis within each scheduling partition to obtain a scheduling scheme within the partition; performing partition constraint verification on the scheduling scheme within the partition to obtain a feasible scheme for the partition; and generating a partitioned scheduling configuration based on the feasible scheme for the partition.

[0078] The dynamic dispatch network is divided into dispatch zones by regional boundary delineation. The topology and node distribution of the dynamic dispatch network are analyzed, and a graph partitioning algorithm is used to divide the network into several relatively independent sub-regions. The partitioning criteria include the geographical location of nodes, electrical connection strength, and functional attributes, prioritizing nodes from the same building or the same power distribution area into the same dispatch zone. A dynamic dispatch network in an industrial park contains 32 resource nodes. 15 nodes in the office building complex are centrally powered through the office area power distribution room and have similar load characteristics, so they are assigned to the office area dispatch zone. 12 equipment nodes and energy storage nodes in the production plant are powered through the industrial power distribution room and have production continuity requirements, so they are assigned to the production area dispatch zone. 5 nodes in supporting facilities such as the canteen and dormitory have relatively independent loads and are assigned to the supporting facilities area dispatch zone. During the partitioning process, the size of each dispatch zone is controlled, with the number of nodes in a single dispatch zone controlled between 8 and 20 to ensure the response speed of dispatch within the zone, and not too few to ensure sufficient self-balancing capacity. The zone boundaries are marked in the dynamic dispatch network, and the nodes on the boundaries serve as inter-zone communication points, responsible for cross-zone energy exchange. The segmentation results are compiled into a scheduling partition list. Each scheduling partition records its partition number, the set of nodes it contains, its boundary nodes, and information on adjacent partitions. The division of scheduling partitions is relatively stable, usually pre-defined according to the physical layout of the campus, and is only re-divided when there are significant topology changes in the dynamic scheduling network. The design of scheduling partitions should ensure good matching between supply and demand within each partition and reduce reliance on cross-partition scheduling.

[0079] An intra-regional supply and demand balance analysis is conducted for each dispatch zone to obtain an intra-regional dispatch plan. For each dispatch zone, the supply capacity and load demand within the zone are statistically analyzed to calculate the intra-regional supply and demand balance status. Supply capacity includes the release capacity of adjustable resources within the dispatch zone and the power supply capacity of the external power grid to the zone; load demand includes the real-time load and predicted load of each energy-consuming device within the dispatch zone. The difference between supply capacity and load demand is calculated; a positive value indicates sufficient supply within the zone, while a negative value indicates a supply-demand gap within the zone. For example, an intra-regional supply and demand balance analysis is conducted for the office area dispatch zone of an industrial park during a summer afternoon. The predicted load demand during this period is 280kW, while the zone's supply capacity is only 200kW, resulting in an 80kW supply-demand gap. It is found that the demand for air conditioning cooling is concentrated on top of the load from lighting and office equipment, while the adjustable resources within the zone only have some air conditioning units with pre-cooling transfer capabilities and some lighting units with dimmable dimming capabilities. An intra-regional dispatch plan needs to be developed to fill the gap. For dispatch zones with supply-demand gaps, an intra-regional resource allocation plan is developed to fill the gap, prioritizing the allocation of adjustable resources within the zone, and selecting resources based on response speed and allocation cost. The resource allocation arrangements for each scheduling partition are compiled into an intra-partition scheduling plan. This plan includes a list of resources to be allocated, the available capacity, the time period for allocation, and the expected results. Each intra-partition scheduling plan is independently developed for each partition, reflecting the principle of partition autonomy. The intra-partition scheduling plan also needs to reserve certain inter-partition scheduling interfaces, allowing requests for support from other scheduling partitions when intra-partition resources are insufficient.

[0080] The intra-area dispatching scheme is validated by zonal constraints to obtain feasible solutions for each zone. The intra-area dispatching scheme for each zone is verified to ensure it meets all operational constraints within that zone. Constraints include power distribution circuit capacity constraints, equipment operational safety constraints, user comfort constraints, and dispatching timing constraints. Each resource allocation arrangement in the intra-area dispatching scheme is iterated to verify whether the circuit load after allocation is within its carrying capacity, whether equipment operating parameters are within allowable ranges, and whether the service quality on the user side meets standards. For example, the intra-area dispatching scheme for the office area of ​​an industrial park includes two measures: 50kW air conditioning pre-cooling and 40kW lighting dimming. Zonal constraint validation of the intra-area dispatching scheme revealed that the 40kW lighting dimming reduced the illuminance in the open office area from 500lx to 280lx, below the national standard's minimum illuminance of 300lx for office spaces. Employees reported that the dim working environment negatively impacted their work efficiency. This allocation arrangement violates user comfort constraints and needs correction. If a scheduling arrangement violates constraints, the intra-zone scheduling plan needs to be modified. This may involve reducing the number of calls or replacing the calling resources. For example, the lighting dimming might be reduced from 40kW to 25kW to maintain an illuminance above 320lx, and an additional 15kW elevator off-peak operation might be added as an alternative to fill the capacity gap. The modified zonal feasible plan is then re-validated until all constraints are met. The intra-zone scheduling plan that passes the constraint validation is considered a zonal feasible plan, which is a safe and executable intra-zone scheduling arrangement. The zonal feasible plan records the list of resources that passed the validation, constraint satisfaction, and safety margin.

[0081] A partitioned scheduling configuration is generated based on the partitioned feasible scheme. The partitioned feasible schemes for each scheduling partition are converted into a standardized configuration format for easy parsing and execution by the partition controller. The configuration format includes four parts: partition identifier, resource list, timing control, and feedback requirements. The partition identifier records the partition number, partition name, and partition boundary information; the resource list records the number, call capacity, and control parameters of each resource; the timing control records the start time, duration, and termination conditions of each resource; and the feedback requirements records the frequency of execution status reporting and exception handling rules. For example, the partitioned feasible scheme for the office area of ​​an industrial park includes three resource call arrangements: air conditioning pre-cooling, lighting dimming, and elevator peak-shaving. When converting the partitioned feasible scheme into a partitioned scheduling configuration, the resource list section is filled with the numbers and call parameters of the three resources. The timing control section sets the timing arrangement as follows: air conditioning pre-cooling starts at 10:00 AM during off-peak hours and lasts for 2 hours; lighting dimming starts at 2:00 PM during peak hours and lasts for 4 hours; and elevator peak-shaving starts before the 5:00 PM rush hour. Fill in the feasible partitioning schemes according to the above format to generate partition scheduling configuration files for each partition. The partition scheduling configurations are stored in JSON format and can be directly loaded and executed by the partition controllers. Each partition scheduling configuration includes a version number and timestamp to ensure that each partition executes the latest configuration. After the partition scheduling configurations are distributed to the partition controllers, the controllers autonomously execute intra-zone scheduling according to the configuration content, without requiring central control of each instruction. The partition scheduling configuration for the production area includes the calling parameters and timing arrangements for energy storage discharge and equipment rotation, while the partition scheduling configuration for the supporting area includes the calling parameters for staggered power consumption in the canteen.

[0082] The system performs global coordination and integration of zone scheduling configurations to output scheduling control commands. It summarizes the zone scheduling configurations of each zone and checks for conflicts or inconsistencies between the scheduling schemes of different zones. When adjacent zones simultaneously request power support from the tie point, support capacity needs to be allocated according to priority. When a zone scheduling scheme causes tie lines to overload, the zone scheme needs to be adjusted or the cross-zone power flow coordinated. For example, in an industrial park, the office zone scheduling configuration is set to reduce the load by 90kW, and the production zone scheduling configuration is set to release 50kW of energy storage. Both zones are scheduled to start simultaneously at 14:00. Global coordination analysis reveals that simultaneous execution of both zones will cause the park's total load to drop sharply by 140kW within 5 minutes, triggering a reverse peak-shaving assessment of the power grid. Therefore, the start times of the two zones need to be staggered: the office zone start time is advanced to 13:50, and the production zone start time is postponed to 14:10, forming a smooth load curve with a 70kW reduction every 10 minutes. The system then performs global optimization of the zone scheduling configurations, aiming to achieve overall supply and demand balance and minimize operating costs for the park, coordinating the scheduling sequence and intensity of each zone. The optimized scheme is converted into executable scheduling and control commands, which include specific control parameters, execution time, and feedback requirements for each resource. These commands are sorted by execution time, forming a time-sequential command queue, and sequentially distributed to each distributed control node. The format of the scheduling and control commands conforms to the park's energy management communication protocol and supports distribution via multiple channels such as industrial Ethernet and wireless networks. After the scheduling and control commands are issued, execution monitoring is initiated, collecting actual response data for each resource and comparing it with the command requirements. Resources with execution deviations exceeding ±10% are subject to closed-loop adjustment.

[0083] To implement the intelligent monitoring and optimized scheduling method for integrated energy in industrial parks corresponding to the above method embodiments, and to achieve the corresponding functions and technical effects. See [link / reference needed]. Figure 2 , Figure 2 This diagram illustrates a structural block diagram of a smart monitoring and optimization scheduling system 200 for integrated energy in a park, as provided in an embodiment of this application. For ease of explanation, only the parts relevant to this embodiment are shown. The smart monitoring and optimization scheduling system 200 for integrated energy in a park, as provided in this embodiment, includes:

[0084] Data acquisition module 201 is used to collect comprehensive energy consumption data and equipment operation status data of the park, and to construct a hierarchical load chain based on the energy consumption data and the equipment operation status data for energy level matching.

[0085] The loss identification module 202 is used to identify energy loss areas by combining the hierarchical load chain and the equipment operating status data, and to classify the energy loss areas according to the loss recoverability to form scheduling advantage areas.

[0086] Load analysis module 203 is used to perform load density analysis on the scheduling advantage area to form a density fluctuation curve, identify a set of transferable loads based on the density fluctuation curve, and perform user-side response potential assessment based on the set of transferable loads to generate a load balancing flow.

[0087] Resource management module 204 is used to perform scheduling capacity analysis on the load balancing flow to extract scheduling redundant resources, perform call conflict analysis on the scheduling redundant resources to obtain resource mutual exclusion relationship identifiers, and form an adjustable load resource pool based on the resource mutual exclusion relationship identifiers according to the response speed.

[0088] The early warning response module 205 is used to detect supply and demand risks for the load balancing flow, obtain energy early warning levels, and trigger the adjustable load resource pool to generate a dynamic scheduling network according to the energy early warning levels.

[0089] The coordination and scheduling module 206 is used to generate partition scheduling configurations based on the dynamic scheduling network, and to globally coordinate and integrate the partition scheduling configurations to output scheduling control commands.

[0090] The aforementioned integrated energy intelligent monitoring and optimization scheduling system 200 for industrial parks can implement the integrated energy intelligent monitoring and optimization scheduling method for industrial parks described in the above-described method embodiments. The options in the above method embodiments are also applicable to this embodiment and will not be detailed here. The remaining contents of this application embodiment can be referred to the contents of the above method embodiments, and will not be repeated in this embodiment.

[0091] The purpose of the above embodiments is to reproduce and derive the technical solution of the present invention by way of example, and to fully describe the technical solution, purpose and effect of the present invention. The purpose is to enable the public to have a more thorough and comprehensive understanding of the disclosure of the present invention, and not to limit the scope of protection of the present invention.

[0092] The above embodiments are not an exhaustive list based on the present invention, and there may be many other embodiments not listed. Any substitutions and improvements made without departing from the concept of the present invention are within the protection scope of the present invention.

Claims

1. A method for intelligent monitoring and optimized scheduling of integrated energy resources in a park, characterized in that, include: Collect comprehensive energy consumption data and equipment operation status data of the park, and construct a hierarchical load chain based on the energy consumption data and the equipment operation status data by matching energy levels. This includes: classifying the equipment operation status data into equipment importance levels to obtain an equipment priority sequence; allocating energy level margin gradients based on the equipment priority sequence to obtain differentiated energy level quotas; matching the energy consumption data and the differentiated energy level quotas to supply and demand to form an energy level matching path; and constructing a hierarchical load chain based on the energy level matching path. The energy loss zone is identified by combining the hierarchical load chain and the equipment operating status data. The energy loss zone is then classified according to its recoverability to form a scheduling advantage region. This includes: analyzing the causes of energy loss in the energy loss zone to identify schedulable losses; evaluating the recovery cost of the schedulable losses to obtain recovery benefit values; prioritizing the recovery benefit values ​​to form a loss control sequence; and delineating the scheduling advantage region based on the loss control sequence. The process involves: analyzing load density in the designated advantageous region to generate a density fluctuation curve; identifying a set of transferable loads based on the density fluctuation curve; and assessing user-side response potential based on the set of transferable loads to generate a load balancing flow. This includes: collecting user response intention data from the set of transferable loads to obtain a response intention level; statistically analyzing the transfer capacity of the set of transferable loads to obtain a load transfer capacity; calculating a comprehensive response potential value based on the response intention level and the load transfer capacity; and generating a load balancing flow based on the comprehensive response potential value. The process involves: performing scheduling capacity analysis on the load balancing flow to extract redundant scheduling resources; and performing call conflict analysis on the redundant scheduling resources to obtain resource mutual exclusion identifiers. This includes: performing device correlation analysis on the redundant scheduling resources to obtain a resource correlation matrix; identifying shared loop resources and independent loop resources based on the resource correlation matrix; performing parallel call constraint detection on the shared loop resources to obtain mutually exclusive resource pairs; generating resource mutual exclusion identifiers based on the mutually exclusive resource pairs and the independent loop resources; and forming an adjustable load resource pool based on the resource mutual exclusion identifiers and tiered by response speed. The process involves: detecting supply and demand risks to obtain energy warning levels for the load balancing flow; triggering multi-level linkage based on the energy warning levels to generate a dynamic scheduling network from the adjustable load resource pool; determining release priorities based on the energy warning levels; extracting target resource groups from the adjustable load resource pool according to the release priorities; performing mutual exclusion checks on the target resource groups to eliminate conflicting resources and form an effective resource set; and constructing a dynamic scheduling network based on the effective resource set. The process of generating partitioned scheduling configurations based on the dynamic scheduling network includes: dividing the dynamic scheduling network into regional boundaries to obtain scheduling partitions; performing supply and demand balance analysis within each scheduling partition to obtain scheduling schemes within the partition; performing partition constraint verification on the scheduling schemes within the partition to obtain feasible partition schemes; generating partitioned scheduling configurations based on the feasible partition schemes; and globally coordinating and integrating the partitioned scheduling configurations to output scheduling control commands.

2. The method according to claim 1, characterized in that, The step of evaluating the recovery cost of the schedulable loss to obtain a recovery benefit value includes: Perform recovery time window analysis on the schedulable loss to obtain the effective recovery period; Resource consumption is estimated during the effective recovery period to obtain the recovery cost; After restoring the schedulable losses, the energy efficiency is estimated to obtain energy savings. The recovery benefit value is calculated based on the energy-saving benefits, the recovery costs, and the effective recovery period.

3. The method according to claim 1, characterized in that, The step of performing parallel call constraint detection on the shared loop resources to obtain mutually exclusive resource pairs includes: The loop capacity is extracted from the shared loop resources to obtain the upper limit of loop capacity. The resource call power is obtained by performing call power statistics on the shared loop resources; The superimposed load value is obtained by superimposing the circuit carrying capacity limit and the resource call power; Based on the superimposed load value, exceedance determination is performed to identify mutually exclusive resource pairs.

4. A comprehensive intelligent monitoring and optimized scheduling system for energy in industrial parks, characterized in that: include: The data acquisition module is used to collect comprehensive energy consumption data and equipment operation status data of the park. Based on the energy consumption data and the equipment operation status data, it performs energy level tiered matching to construct a hierarchical load chain, including: classifying the equipment operation status data into equipment importance levels to obtain an equipment priority sequence; allocating energy level margin gradients based on the equipment priority sequence to obtain differentiated energy level quotas; matching the energy consumption data and the differentiated energy level quotas to supply and demand to form an energy level matching path; and constructing a hierarchical load chain based on the energy level matching path. The loss identification module is used to identify energy loss areas by combining the hierarchical load chain and the equipment operating status data, and to classify the energy loss areas according to their recoverability to form scheduling advantage areas. This includes: analyzing the causes of loss in the energy loss areas to identify schedulable losses; evaluating the recovery costs of the schedulable losses to obtain recovery benefit values; prioritizing the recovery benefit values ​​to form a loss control sequence; and delineating scheduling advantage areas based on the loss control sequence. The load analysis module is used to perform load density analysis on the scheduling advantage area to form a density fluctuation curve, identify a set of transferable loads based on the density fluctuation curve, and evaluate the user-side response potential based on the set of transferable loads to generate a load balancing flow. This includes: collecting user response intention data from the set of transferable loads to obtain a response intention level; performing transfer capacity statistics on the set of transferable loads to obtain a load transfer capacity; calculating a comprehensive response potential value based on the response intention level and the load transfer capacity; and generating a load balancing flow based on the comprehensive response potential value. The resource management module is used to perform scheduling capacity analysis on the load balancing flow to extract scheduling redundant resources, and to perform call conflict analysis on the scheduling redundant resources to obtain resource mutual exclusion relationship identifiers. This includes: performing device correlation analysis on the scheduling redundant resources to obtain a resource correlation matrix; identifying shared loop resources and independent loop resources based on the resource correlation matrix; performing parallel call constraint detection on the shared loop resources to obtain mutually exclusive resource pairs; generating resource mutual exclusion relationship identifiers based on the mutually exclusive resource pairs and the independent loop resources; and forming an adjustable load resource pool based on the resource mutual exclusion relationship identifiers and tiered by response speed. The early warning response module is used to detect supply and demand risks for the load balancing flow to obtain energy early warning levels, and to trigger the adjustable load resource pool to generate a dynamic scheduling network according to the energy early warning levels through multi-level linkage. This includes: determining release priorities based on the energy early warning levels; extracting target resource groups from the adjustable load resource pool according to the release priorities; performing mutual exclusion checks on the target resource groups to eliminate conflicting resources and form an effective resource set; and constructing a dynamic scheduling network based on the effective resource set. The coordination and scheduling module is used to generate partitioned scheduling configurations based on the dynamic scheduling network, including: dividing the dynamic scheduling network into regional boundaries to obtain scheduling partitions; performing supply and demand balance analysis within each scheduling partition to obtain intra-partition scheduling schemes; performing partition constraint verification on the intra-partition scheduling schemes to obtain partitioned feasible schemes; generating partitioned scheduling configurations based on the partitioned feasible schemes; and performing global coordination and integration of the partitioned scheduling configurations to output scheduling control commands.

Images