Enterprise energy intelligent optimization method and system based on big data analysis

Abstract

The invention discloses an enterprise energy intelligent optimization method and system based on big data analysis, and relates to the technical field of energy optimization. Performing graph structure expression on the energy consumption behaviors among the process nodes by using the process behavior data graph XW; monitoring the energy consumption level of each process node in each time window, taking any two process nodes in the same time window as a combination pair, and analyzing the coupling energy consumption superimposed effect of all the combination pairs in each time window; analyzing the conflict degree of energy consumption superposition in the operation process of the plurality of process nodes in each time window, obtaining the conflict degree index Zct of each time window, then judging whether the energy consumption superposition conflict in the operation process of the plurality of process nodes in the corresponding time window has an overload risk or not, and executing a corresponding energy optimization strategy.

Description

Technical Field

[0001] This invention relates to the field of energy optimization technology, specifically to a method and system for intelligent energy optimization in enterprises based on big data analysis. Background Technology

[0002] With the rapid development of artificial intelligence, the Internet of Things, and edge computing, big data analytics has been widely applied to various industries, including manufacturing, transportation, finance, and healthcare, becoming a core engine driving the intelligent, adaptive, and efficient operation of systems. In the industrial sector, especially in manufacturing, big data analytics is not only used for predictive maintenance, quality traceability, and capacity assessment, but has also been extended to energy optimization management, a fundamental support layer for enterprise operations. In the field of enterprise energy optimization, in such complex manufacturing systems, energy optimization management of precision electronic manufacturing production lines, as one of the key application directions of big data analytics, is gradually shifting from static statistics of energy consumption to predictive control and intelligent optimization based on dynamic behavioral data.

[0003] In current precision electronics manufacturing processes, most companies still employ traditional strategies such as static scheduling, threshold alarms, or experience-based peak shaving for scheduling multiple parallel or overlapping process nodes in high-energy-consumption scenarios. These methods generally lack accurate identification of dynamic changes in process behavior and struggle to perceive the energy consumption coupling characteristics between different process nodes within the same time window. For example, when the reflow soldering process and the wafer packaging cooling process run simultaneously during a certain time period, although the two process devices are compliant individually, their combined load exceeds the local energy supply capacity threshold, creating an overload risk. Furthermore, there is a lack of early warning capabilities for such non-single-point anomalies and cross-node superposition conflicts, making it difficult to perform graphical modeling of coupled energy consumption and establishing a coupling risk control mechanism between cross-node scheduling strategies. This leaves many scheduling behaviors in a reactive state, leading to production cycle interruptions and energy allocation delays, significantly reducing system stability and energy efficiency. Therefore, there is an urgent need to construct an energy optimization management mechanism based on big data analysis and centered on a time-window energy consumption coupling graph to achieve proactive perception and adaptive scheduling of parallel process conflicts. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a method and system for intelligent energy optimization for enterprises based on big data analysis, which solves the problems mentioned in the background technology.

[0005] To achieve the above objectives, the present invention provides the following technical solution: an enterprise energy intelligent optimization method based on big data analysis, comprising the following steps: S1. During the operation of multiple process nodes within the precision electronics manufacturing production line of an enterprise, the energy consumption behavior between each process node is expressed in a graph structure using the process behavior data map XW, and each process node in the same time window in the map is uniformly marked. S2. Monitor the energy consumption level of each process node within each time window, and take any two process nodes within the same time window as a pair, and analyze the superposition effect of the coupling energy consumption of all pairs within each time window. S3. Based on the maximum power that the energy supply system of the enterprise's production line can stably output within a unit time window, analyze the degree of conflict of energy consumption superposition during the operation of multiple process nodes within each time window, and obtain the conflict degree index Zct for each time window. S4. Based on the conflict degree index Zct value of each time window, determine whether there is an overload risk of energy consumption superposition conflict during the operation of multiple process nodes within the corresponding time window, and execute the corresponding energy optimization strategy.

[0006] Preferably, step S1 specifically includes: S11. During the operation of multiple process nodes within the enterprise's precision electronics manufacturing production line, based on IoT architecture, edge computing, and high-frequency sensing devices, the concurrent start count, real-time high-power trigger point, and process chain synchronization lag time of each process node within each time window are obtained. The specific process includes: S111. By using the event trigger logs and PLC instruction streams embedded in the equipment control system, identify the concurrent startup behavior of multiple process nodes within a unit time window, perform real-time statistics on the trigger frequency of equipment state changes within a unit time window, form statistics and record them, and obtain the concurrent startup count Cqd of each process node within each time window. S112. Based on the high-frequency energy consumption metering device integrated in the process control system, the energy consumption equipment of each process node is sampled in real time, and the energy consumption derivative threshold within the sliding window is set. The trigger point of the power change in the sampled signal within the unit time window is identified. The edge computing unit completes the preliminary judgment and uploads it to the cloud for processing to obtain the real-time high power trigger point Ngg of each process node within each time window. S113. Based on the start and finish timestamps of each process, the interval between task delivery and task receipt is automatically recorded each time a product is transferred from the upstream process node to the downstream process node. The average delay is extracted based on statistical rules to form the process chain synchronization lag time Tzh of each process node within each time window.

[0007] Preferably, step S1 further includes: S12. Based on the concurrent start-up count Cqd of each process node in each time window, the real-time high-power trigger point Ngg, and the synchronization lag time Tzh of the process chain, a sliding window matching algorithm is introduced to reorganize the timing and form a process behavior data map XW that scrolls in a sliding window manner, which is used as the input data of the energy consumption coupling analysis model. S13. Based on the constructed process behavior data map XW, the energy consumption behavior between each process node is expressed in a graph structure, and each process node in the same time window in the graph is uniformly marked.

[0008] Preferably, step S2 specifically includes: S21. After uniformly marking each process node based on the same time window, the energy consumption level of each process node in each time window is monitored by the high-frequency energy consumption metering device integrated in the process control system, and the energy consumption value Enh of each process node in each time window is obtained. The high-frequency energy consumption metering device includes a three-phase smart energy meter, a power quality analyzer, a heat mass flow meter and an edge acquisition device.

[0009] Preferably, step S2 further includes: S22. Let any two process nodes within the same time window be denoted as a pair (i, j). Based on the energy consumption value Enh of each process node within each time window obtained in step S11, and combined with the time series similarity measurement algorithm, analyze the degree of resource usage overlap between any two process nodes within the same time window, and obtain the resource usage overlap coefficient Xcd of each pair within each time window. The specific formula is as follows: ; In the formula, It represents the resource usage overlap coefficient between the i-th and j-th process nodes within the k-th time window. This represents the energy consumption value at the i-th process node within the k-th time window. Let be the energy consumption value of the j-th process node within the k-th time window, where It represents the minimum value of the intersection of the actual energy consumption of the i-th process node and the j-th process node within the k-th time window. It is represented as the maximum value of the union of energy consumption of the i-th process node and the j-th process node within the k-th time window.

[0010] Preferably, step S2 further includes: S23. Based on the timestamp recording function in the process control system, and combined with the resource usage overlap coefficient Xcd of each pair in each time window obtained in step S22, record the operation end timestamp and operation start timestamp of the two process nodes in the corresponding pair in the corresponding time window, calculate the operation time difference of the two process nodes in each pair, and record it as the execution operation time difference Tsc of each pair in each time window. S24. Correlate the resource usage overlap coefficient Xcd of each pair within each time window with the corresponding execution operation time difference Tsc. After dimensionless processing, analyze the energy consumption coupling degree of any two process nodes executing operations within the same time window, and obtain the execution energy consumption coupling coefficient Xoh of each pair within each time window. Specifically, it is obtained through the following formula: ; In the formula, It represents the execution energy coupling coefficient of the pair of process nodes i and j within the k-th time window. It represents the resource usage overlap coefficient between the i-th and j-th process nodes within the k-th time window. Let $\mathbf{i}$ be the execution time difference between the pair of process nodes $i$ and $j$ within the $k$-th time window, where $\mathbf{i}$ is the execution time difference between the pair of process nodes $i$ and $j$. This is expressed as the operation time difference adjustment factor. It is represented as an exponential function.

[0011] Preferably, step S2 further includes: S25. Correlate the execution energy consumption coupling coefficient Xoh of each pair within each time window obtained in step S24 with the energy consumption value Enh of each process node within each time window obtained in step S11, analyze the coupling energy consumption superposition effect of all pairs within each time window, and obtain the coupling energy consumption superposition value Zdj of each time window, which is specifically obtained through the following formula: ; In the formula, This is represented as the sum of the coupling energy consumption in the k-th time window. This represents the energy consumption value at the i-th process node within the k-th time window. This is represented as the energy consumption value at the j-th process node within the k-th time window. Let be the execution energy coupling coefficient of the pair of process nodes i and j within the k-th time window, where This represents the number of pairings between any two process nodes within the same time window.

[0012] Preferably, step S3 specifically includes: S31. The maximum power that the energy supply system of the enterprise's production line can stably output within a unit time window is recorded as the load threshold Fz. It is correlated with the coupled energy consumption superposition value Zdj of each time window obtained in step S25. After dimensionless processing, the degree of conflict of energy consumption superposition during the operation of multiple process nodes within each time window is analyzed to obtain the conflict degree index Zct of each time window, which is obtained by the following formula: ; In the formula, This is represented as the conflict level index for the k-th time window. It is represented as the superposition value of the coupled energy consumption of the k-th time window.

[0013] Preferably, step S4 specifically includes: S41. Based on the conflict degree index Zct values ​​obtained in step S31 for each time window, determine whether there is an overload risk due to the superposition of energy consumption conflicts during the operation of multiple process nodes within the corresponding time window, and execute the corresponding energy optimization strategy. The specific content is as follows: If the conflict index Zct of the corresponding time window is greater than 1, it means that the maximum power that the energy supply system can stably output within the current time window cannot meet the operation of multiple process nodes. In this case, it is judged that there is an overload risk in the operation of multiple process nodes within the corresponding time window. At this time, the time window corresponding to the conflict index Zct is recorded as the overload time window, and an energy optimization command is issued to the outside. If the conflict index Zct of the corresponding time window is less than or equal to 1, it means that the maximum power that the energy supply system can stably output within the current time window can meet the operation of multiple process nodes. Therefore, it is determined that there is no risk of overload in the operation of multiple process nodes within the corresponding time window. The time window corresponding to the conflict index Zct is recorded as the normal time window set. When all time windows are marked as normal time windows, a normal operation signal is sent out. S42. Upon receiving the energy optimization instruction, based on the process behavior data map... As input data for the energy consumption coupling analysis model, combined with heuristic search algorithm, local disturbance rearrangement mechanism and conflict sensitivity weight adjustment method, the concurrent start-up count of the corresponding process node in the overload time window, the real-time high power trigger point and the process chain synchronization lag time are updated and adjusted, which are used as new input data for the energy consumption coupling analysis model. Steps S2 and S3 are repeated until a normal operation signal is issued.

[0014] The enterprise energy intelligent optimization system based on big data analysis includes a graph construction module, an overlay analysis module, a conflict analysis module, and an energy optimization module. The graph construction module uses the process behavior data graph XW to represent the energy consumption behavior between various process nodes during the operation of multiple process nodes within the enterprise's precision electronics manufacturing production line, and uniformly marks each process node in the same time window in the graph. The superposition analysis module monitors the energy consumption level of each process node within each time window. Any two process nodes within the same time window are recorded as a pair, and the superposition effect of the coupling energy consumption of all pairs within each time window is analyzed. The conflict analysis module analyzes the degree of conflict of energy consumption superposition during the operation of multiple process nodes in each time window based on the maximum power that the energy supply system of the enterprise production line can stably output within a unit time window, and obtains the conflict degree index Zct for each time window. The energy optimization module determines whether there is an overload risk due to the superposition of energy consumption conflicts during the operation of multiple process nodes within the corresponding time window based on the value of the conflict degree index Zct for each time window, and then executes the corresponding energy optimization strategy.

[0015] This invention provides a method and system for intelligent energy optimization in enterprises based on big data analysis, which has the following beneficial effects: (1) In the complex scenario of multiple process nodes operating in parallel on a precision electronic manufacturing production line, based on continuous perception and structural modeling of process behavior data, high-precision identification and conflict prediction of energy consumption coupling relationship of process nodes within the same time window can be achieved, thereby enabling early judgment of overload risk and dynamic adjustment of cross-node scheduling strategy. Compared with the existing traditional energy management method that relies on experience-based peak shifting and static energy consumption statistics, it significantly improves the modeling ability of nonlinear cross-energy consumption behavior and overload response speed, effectively reduces scheduling lag, overload triggering and fluctuation risks, and improves the foresight, accuracy and intelligence level of enterprise energy optimization management.

[0016] (2) By collecting and reconstructing the concurrent start-up count Cqd, real-time high-power trigger point Ngg, and process chain synchronization lag time Tzh of each process node based on edge computing and high-frequency sampling technology, and introducing a sliding window matching algorithm to construct a rolling updated process behavior data map XW, this invention can accurately restore the behavior rhythm and load status of multiple process nodes within a unit time window, realizing the transformation of process behavior from discrete recording to structured expression. Compared with the traditional coarse-grained data method that only records the start and end times of the process, the process behavior data map XW construction mechanism supports unified time-series analysis and graph model input of cross-node behavior sequences, providing a high-timeliness and high-resolution data foundation for subsequent energy consumption coupling relationship modeling and dynamic control, significantly enhancing the system's cognitive dimension and modeling depth of complex production processes.

[0017] (3) By forming a pair of any two process nodes within the same time window, and combining the energy consumption value Enh, resource usage overlap coefficient Xcd, and execution operation time difference Tsc of each node, the execution energy consumption coupling coefficient Xoh is constructed, and the coupling energy consumption superposition value Zdj is further obtained, thus realizing the quantitative analysis of the multi-dimensional energy consumption coupling effect between process nodes. This mechanism breaks through the limitations of traditional single-node energy consumption assessment, and can identify the cross superposition effect of multiple processes running concurrently in the high energy consumption stage on the same energy supply path, accurately reveal potential coupling conflicts, and effectively support the pre-judgment of structural identification and scheduling intervention of parallel process load coupling.

[0018] (4) Based on the superimposed value of coupled energy consumption Zdj in each time window, the load threshold Fz of the enterprise energy system is introduced, the conflict degree index Zct in each time window is calculated, and the time window is divided into overload time window and normal time window through the judgment mechanism, so as to realize the time-segmented dynamic identification and marking of potential energy consumption conflicts in the concurrent operation of multiple processes; based on the process behavior data map XW, the process nodes marked as overload time windows are subjected to heuristic search and local disturbance optimization strategies, and their concurrent start-up parameters Cqd, high power triggering behavior Ngg and synchronous lag characteristics Tzh are adaptively adjusted to construct a cross-time window coupling peak shaving scheme, effectively alleviate the risk of concentrated energy consumption operation, avoid sudden overload, and ensure the continuity of production cycle and the stable operation of energy supply system. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the enterprise energy intelligent optimization method based on big data analysis according to the present invention. Figure 2 This is a block diagram of the enterprise energy intelligent optimization system based on big data analysis according to the present invention; Figure 3 This is a logic diagram for step S1 of the present invention; Figure 4 This is a logic diagram for step S2 of the present invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Example 1 Please see Figure 1 This invention provides a smart energy optimization method for enterprises based on big data analysis, comprising the following steps: S1. During the operation of multiple process nodes within the precision electronics manufacturing production line of an enterprise, the energy consumption behavior between each process node is expressed in a graph structure using the process behavior data map XW, and each process node in the same time window in the map is uniformly marked. S2. Monitor the energy consumption level of each process node within each time window, and take any two process nodes within the same time window as a pair, and analyze the superposition effect of the coupling energy consumption of all pairs within each time window. S3. Based on the maximum power that the energy supply system of the enterprise's production line can stably output within a unit time window, analyze the degree of conflict of energy consumption superposition during the operation of multiple process nodes within each time window, and obtain the conflict degree index Zct for each time window. S4. Based on the conflict degree index Zct value of each time window, determine whether there is an overload risk of energy consumption superposition conflict during the operation of multiple process nodes within the corresponding time window, and execute the corresponding energy optimization strategy.

[0022] In this embodiment, addressing the complex scenario of multiple process nodes operating in parallel within a precision electronics manufacturing production line, a time-series rolling process behavior data graph (XW) is constructed. This achieves a graph-structured representation of the energy consumption behavior of process nodes and a unified time window labeling, thus providing a precise data foundation for subsequent energy consumption coupling modeling and dynamic control. By introducing a pairwise analysis mechanism, this method can uncover resource usage overlaps and energy consumption behavior intersections between process nodes within the same time window, thereby quantifying and identifying the coupled energy consumption superposition effect. Furthermore, by combining the stable load threshold of the enterprise's energy system, the conflict degree index (Zct) within each time window is analyzed to achieve cross-… This method proactively assesses the overload risks arising from centralized energy consumption at nodes. Unlike traditional passive response modes based on static statistics and threshold alarms, this method, driven by data, graphical representation, and coupled analysis, dynamically adapts to changes in process behavior, accurately identifies nonlinear cross-energy consumption conflicts in concurrent operation of multiple nodes, and automatically triggers energy optimization strategies based on real-time conflict identification results, achieving linkage matching between process scheduling rhythm and energy load capacity. Especially in manufacturing environments with frequent parallel operation of complex process chains, this method significantly improves the accuracy and response speed of energy scheduling, providing a new optimization path for achieving efficient, safe, and intelligent operation of enterprise-level energy.

[0023] Example 2 Please refer to Figure 1 and Figure 3 Specifically, the steps in S1 include: S11. During the operation of multiple process nodes within the enterprise's precision electronics manufacturing production line, based on IoT architecture, edge computing, and high-frequency sensing devices, the concurrent start count, real-time high-power trigger point, and process chain synchronization lag time of each process node within each time window are obtained. The specific process includes: S111. By using the event trigger logs and PLC instruction streams embedded in the equipment control system, identify the concurrent startup behavior of multiple process nodes within a unit time window, perform real-time statistics on the trigger frequency of equipment state changes within a unit time window, form statistics and record them, and obtain the concurrent startup count Cqd of each process node within each time window. S112. Based on the high-frequency energy consumption metering device integrated in the process control system, the energy consumption equipment of each process node is sampled in real time, and the energy consumption derivative threshold within the sliding window is set. The trigger point of the power change in the sampled signal within the unit time window is identified. The edge computing unit completes the preliminary judgment and uploads it to the cloud for processing to obtain the real-time high power trigger point Ngg of each process node within each time window. S113. Based on the start and finish timestamps of each process, the interval between task delivery and task receipt is automatically recorded each time a product is transferred from the upstream process node to the downstream process node. The average delay is extracted based on statistical rules to form the process chain synchronization lag time Tzh of each process node within each time window.

[0024] It should be noted that the concurrent start count (Cqd), real-time high-power trigger point (Ngg), and process chain synchronization lag time (Tzh) of each process node within each time window are key dynamic behavioral parameters reflecting the operational characteristics of process nodes in a precision electronics manufacturing production line. Specifically, the concurrent start count (Cqd) refers to the number of times multiple process devices start synchronously within the same time window. This can be obtained by analyzing the event trigger logs and PLC instruction streams embedded in the equipment control system to statistically determine the frequency of overlap between the idle and running states of multiple nodes within a short period. The real-time high-power trigger point (Ngg) of each process node within each time window represents the number of times the equipment energy consumption curve appears within a unit time window. At the moment of power surge, high-frequency energy consumption sampling data is used to set a sliding window to detect the power derivative. If it exceeds the set threshold, it is marked as a trigger point and reported by the edge computing unit. The process chain synchronization lag time Tzh of each process node in each time window reflects the average time delay generated between upstream and downstream nodes in the process of material delivery and receipt. Based on the start and end timestamp records of each process in the process control system, the time difference between the completion and acceptance of tasks of upstream and downstream nodes is calculated and then normalized according to statistical rules. The three together constitute the dynamic load and timing characteristics of process nodes in the time window, which is the basis for subsequent construction of process behavior map, mining of energy consumption coupling relationship, and prediction of scheduling conflict.

[0025] Specifically, the S1 steps also include: S12. Based on the concurrent start-up count Cqd of each process node in each time window, the real-time high-power trigger point Ngg, and the synchronization lag time Tzh of the process chain, a sliding window matching algorithm is introduced to reorganize the timing and form a process behavior data map XW that scrolls in a sliding window manner, which is used as the input data of the energy consumption coupling analysis model. S13. Based on the constructed process behavior data map XW, the energy consumption behavior between each process node is expressed in a graph structure, and each process node in the same time window in the graph is uniformly marked.

[0026] It should be noted that the Process Behavior Data Graph XW is a dynamic energy consumption characteristic model expressed in graph structure. It is used to characterize the behavior and interrelationships of multiple process nodes in a precision electronics manufacturing production line within different time windows. Essentially, it characterizes the operational characteristics of each process node within a specific time window as node attributes, and establishes connection edges between nodes based on operation sequence and coupling characteristics within a sliding time window, forming a time-driven, state-related graph structure. It can dynamically record the energy consumption coupling trends and conflict potentials between process nodes, serving as the core input for subsequent coupled energy consumption analysis and conflict assessment models. This enables panoramic modeling of energy consumption behavior and accurate identification of local coupling hotspots, thereby providing a structured decision-making basis for adaptive optimization of energy scheduling strategies.

[0027] In this embodiment, the process behavior data map construction mechanism proposed in step S1, based on IoT architecture, edge computing, and high-frequency sensing technology, proposes a dynamically responsive energy consumption behavior modeling method around the operating characteristics of process nodes, significantly improving the ability to identify the timing of complex process states. By acquiring the concurrent start count Cqd, real-time high-power trigger point Ngg, and process chain synchronization lag time Tzh within each time window, this method can not only capture the collaborative relationship of process actions from the equipment instruction layer, but also accurately locate the trigger time of load peaks from energy consumption mutations, and quantify the coordination delay between upstream and downstream processes from product flow behavior, truly realizing multi-dimensional process state modeling from process operation characteristics, energy consumption behavior, and timing coordination. Furthermore, by introducing a sliding window matching algorithm, the acquired timing behavior data is dynamically reorganized, and a rolling updated process behavior map is constructed. The XW process behavior data graph not only preserves the temporal evolution information of process states but also endows the data model with the ability to continuously adapt to behavioral abrupt changes and frequent cross-operations, enhancing the system's ability to express nonlinear, short-cycle, and multi-node process states. Especially in electronic manufacturing scenarios where multiple key devices frequently operate concurrently and multiple process tasks are intertwined, this graphing mechanism can significantly improve the system's ability to visualize and predict energy consumption behavior trends, providing a high-fidelity and high-precision data input foundation for subsequent coupling analysis and conflict identification, thus laying the core support conditions for the operation of the entire energy optimization model. Compared with traditional energy consumption assessment frameworks based on single-node energy consumption logs or fixed-time sampling methods, the graphing expression structure constructed in this invention has stronger dynamic adaptability, structural clarity, and temporal continuity, significantly improving the energy optimization management's insight into production behavior and response efficiency.

[0028] Example 3 Please refer to Figure 1 and Figure 4 Specifically, the steps in S2 include: S21. After uniformly marking each process node based on the same time window, the energy consumption level of each process node in each time window is monitored by the high-frequency energy consumption metering device integrated in the process control system, and the energy consumption value Enh of each process node in each time window is obtained. The high-frequency energy consumption metering device includes a three-phase smart energy meter, a power quality analyzer, a heat mass flow meter and an edge acquisition device.

[0029] It should be noted that the energy consumption value Enh refers to the comprehensive energy quantification index actually consumed by a process node during its operation, usually expressed in equivalent energy consumption units of electrical and thermal energy. It is a fundamental parameter for measuring the energy consumption intensity of each process node and a core basis for subsequent assessment of energy consumption coupling relationships and conflict risks between nodes. In the actual acquisition process, high-frequency energy consumption metering equipment such as three-phase smart energy meters, power quality analyzers, heat and mass flow meters, and edge acquisition devices deployed on process equipment are used to sample the operating energy consumption of nodes within a unit time window at high resolution. After sliding window integration, noise filtering, and time window summation, a standardized energy consumption value Enh is output. It not only reflects the energy consumption level of the process itself, but also provides data support for identifying high energy consumption behavior of nodes, overlapping use of computing resources, and constructing coupling relationships.

[0030] Specifically, the S2 steps also include: S22. Let any two process nodes within the same time window be denoted as a pair (i, j). Based on the energy consumption value Enh of each process node within each time window obtained in step S11, and combined with the time series similarity measurement algorithm, analyze the degree of resource usage overlap between any two process nodes within the same time window, and obtain the resource usage overlap coefficient Xcd of each pair within each time window. The specific formula is as follows: ; In the formula, It represents the resource usage overlap coefficient between the i-th and j-th process nodes within the k-th time window. This represents the energy consumption value at the i-th process node within the k-th time window. Let be the energy consumption value of the j-th process node within the k-th time window, where It represents the minimum value of the intersection of the actual energy consumption of the i-th process node and the j-th process node within the k-th time window. It is represented as the maximum value of the union of energy consumption of the i-th process node and the j-th process node within the k-th time window.

[0031] It should be noted that the formula in step S22 logically addresses the technical bottleneck raised in the background technology—the difficulty in perceiving the energy consumption coupling characteristics of different process nodes within the same time window. The background technology points out that traditional methods only statically monitor the energy consumption of a single node, making it difficult to identify the superimposed load effects formed by multiple parallel or overlapping processes within the same time period, resulting in a lack of early awareness of cross-node energy consumption conflicts. This formula, however, normalizes and compares the energy consumption values ​​of any two process nodes within the same time window, constructing a standardized resource usage overlap coefficient Xcd based on the ratio of the minimum to the maximum value. This allows for the assessment of energy consumption without relying on absolute power thresholds. The system determines whether there is significant overlap in resource usage between nodes. When the resource usage overlap coefficient Xcd is close to 1, it indicates that the two process nodes are operating at the same energy consumption level and have high overlap. Conversely, it indicates that their energy consumption behaviors do not affect each other or are staggered. Its role is to serve as the core input variable in the entire energy consumption coupling analysis model, supporting further analysis of execution coupling and energy consumption superposition status, and ultimately used for conflict intensity assessment and optimized scheduling decisions. Therefore, the resource usage overlap coefficient Xcd is a fundamental parameter for realizing the transition from single-point energy consumption monitoring to multi-node collaborative coupling perception, and is a key quantitative tool for breaking the rigidity of traditional energy consumption scheduling and improving perception granularity and forward-looking control capabilities.

[0032] Specifically, the S2 steps also include: S23. Based on the timestamp recording function in the process control system, and combined with the resource usage overlap coefficient Xcd of each pair in each time window obtained in step S22, record the operation end timestamp and operation start timestamp of the two process nodes in the corresponding pair in the corresponding time window, calculate the operation time difference of the two process nodes in each pair, and record it as the execution operation time difference Tsc of each pair in each time window. It should be noted that the execution operation time difference Tsc refers to the time difference between the start and end of the operation task of any two process nodes in a pair of coupling (i, j) within the same time window. It reflects the actual time misalignment of the two process nodes when executing the process task. Its function is to reveal the synchronous or lagging relationship between the two nodes in the operation execution. Combined with the resource usage overlap coefficient Xcd, it helps to assess the timing impact of energy consumption superposition between nodes, that is, whether there is a risk of local load peaks caused by high coupling and simultaneous execution. S24. Correlate the resource usage overlap coefficient Xcd of each pair within each time window with the corresponding execution operation time difference Tsc. After dimensionless processing, analyze the energy consumption coupling degree of any two process nodes executing operations within the same time window, and obtain the execution energy consumption coupling coefficient Xoh of each pair within each time window. Specifically, it is obtained through the following formula: ; In the formula, It represents the execution energy coupling coefficient of the pair of process nodes i and j within the k-th time window. It represents the resource usage overlap coefficient between the i-th and j-th process nodes within the k-th time window. Let $\mathbf{i}$ be the execution time difference between the pair of process nodes $i$ and $j$ within the $k$-th time window, where $\mathbf{i}$ is the execution time difference between the pair of process nodes $i$ and $j$. This is represented as the operation time difference adjustment factor, which is obtained based on a combination of historical operation data analysis and system sensitivity calibration. It is represented as an exponential function.

[0033] It should be noted that the formula in step S24 reflects the strength of the synergy between the energy consumption superposition behavior of two process nodes within the same time window. This is a response to the difficulty in accurately identifying the risk of coupling conflict between process nodes within the same time window in the background technology. The background technology points out that traditional methods often ignore the time difference and concurrent behavior between different nodes, relying only on static alarm thresholds or single-device energy consumption limits, making it difficult to identify energy consumption conflicts formed by cross-node superposition in actual operation. This formula not only considers the similarity of resource usage between the two nodes, but also introduces the adjustment effect of the operation execution time difference Tsc. When the operation times of the two nodes are close, the exponential term approaches 1, and the coupling strength is the greatest. Conversely, if the operation interval is large, the coupling coefficient decreases exponentially with the time difference, reflecting the natural decay characteristic of its energy consumption coupling in time. The role of the execution energy consumption coupling coefficient Xoh is to comprehensively integrate the two dimensions of spatial resource overlap and time-coordinated execution, thereby realizing the real risk modeling and early warning of cross-node concentrated energy consumption operation, filling the technical gap of traditional methods in dynamic control of coupling strength and perception of behavioral coordination, and improving the sensitivity identification and intelligent scheduling capabilities of complex multi-node load states.

[0034] Specifically, the S2 steps also include: S25. Correlate the execution energy consumption coupling coefficient Xoh of each pair within each time window obtained in step S24 with the energy consumption value Enh of each process node within each time window obtained in step S11, analyze the coupling energy consumption superposition effect of all pairs within each time window, and obtain the coupling energy consumption superposition value Zdj of each time window, which is specifically obtained through the following formula: ; In the formula, This is represented as the sum of the coupling energy consumption in the k-th time window. This represents the energy consumption value at the i-th process node within the k-th time window. This is represented as the energy consumption value at the j-th process node within the k-th time window. Let be the execution energy coupling coefficient of the pair of process nodes i and j within the k-th time window, where This represents the number of pairings between any two process nodes within the same time window.

[0035] It should be noted that the formula in step S25 constructs a comprehensive index to measure the overall coupling energy consumption intensity of the current time window. This addresses the core problem pointed out in the background technology: the difficulty in identifying the systemic load risk caused by the superimposed energy consumption of multiple parallel process nodes in the same time window. The background technology mentions that traditional methods are difficult to identify potential local overload or conflict risks from the interaction behavior between multiple process nodes, and still rely on a coarse scheduling strategy based on the static load of a single node. However, this formula analyzes the coupling behavior of all process nodes in the same time window pairwise, comprehensively considering the energy consumption value of the node itself and the resource coupling strength of the process nodes in the same time window. It effectively quantifies the superimposed coupling energy consumption value Zdj caused by all node combinations, forming a global index with actual risk identification capability. It is an important basic quantity for subsequent judgment on whether there is superimposed energy consumption conflict.

[0036] In this embodiment, step S2 focuses on the identification and analysis of the superposition effect of coupled energy consumption. It breaks through the traditional analysis paradigm of energy consumption statistics based on a single node, realizing the identification, modeling, and quantitative calculation of coupled energy consumption for process node pairs within the same time window. This step obtains the energy consumption value Enh of each process node at a micro-time scale using a high-frequency energy consumption metering device, improving response timeliness while ensuring data accuracy. Next, any two process nodes operating within the same time window are combined into a pair (i, j). Using a time series similarity measurement algorithm, their resource usage overlap coefficient Xcd is calculated, truly reflecting the degree of energy resource sharing between the two process nodes within the same time window, thus characterizing the coupling density of energy resources in the spatial dimension. Based on this, the operational behavior characteristics of each pair (i, j) are further explored. Using the timestamp data recorded in the process control system, their execution operation time difference Tsc is calculated, reflecting the collaborative or conflicting relationships of each process node in the time dimension. Through nonlinear combination and indexing... By weighting the data, the execution energy consumption coupling coefficient Xoh is finally calculated. This not only captures the dual effects of spatial coupling and temporal coordination, but also has the ability to dynamically respond to changes in the production scheduling rhythm of different nodes. This step is not limited to the stacking of static total energy consumption, but integrates the spatiotemporal characteristics of energy consumption behavior, realizing high-granularity and structured energy consumption interaction modeling. Finally, the calculation of the coupled energy consumption superposition value Zdj based on the execution energy consumption coupling coefficient Xoh and the energy consumption value Enh of each pair in each time window provides highly reliable quantitative support for subsequent energy consumption conflict judgment and optimization strategy triggering. Compared with the traditional energy consumption analysis method that only focuses on the load of a single device, it can more comprehensively depict the complex energy consumption interaction relationship between nodes, identify local high-risk coupling areas in advance, and avoid misjudgment and energy consumption peak bursts caused by ignoring the coupling of multiple nodes. This provides a more scientific, forward-looking and efficient energy consumption control foundation for precision electronic manufacturing production lines. Its advantages are: it not only has the ability to identify energy consumption anomalies of the nodes themselves, but also has the ability to judge coupling conflicts brought about by the interaction between nodes.

[0037] Example 4 Please refer to Figure 1 Specifically, the S3 steps include: S31. The maximum power that the energy supply system of the enterprise's production line can stably output within a unit time window is recorded as the load threshold Fz. It is correlated with the coupled energy consumption superposition value Zdj of each time window obtained in step S25. After dimensionless processing, the degree of conflict of energy consumption superposition during the operation of multiple process nodes within each time window is analyzed to obtain the conflict degree index Zct of each time window, which is obtained by the following formula: ; In the formula, This is represented as the conflict level index for the k-th time window. It is represented as the superposition value of the coupled energy consumption of the k-th time window.

[0038] It should be noted that the formula in step S31 directly addresses the key shortcoming of the background technology: the difficulty in determining whether the concentrated energy consumption behavior of multiple process nodes in the same time window exceeds the system's carrying capacity. Traditional methods are often based on single-point load of equipment or static alarm rules, making it difficult to identify the superimposed load caused by multiple parallel processes overlapping in time, thus missing timely warnings of overload risks. By analyzing the behavior of all process nodes within the time window to obtain the coupled energy consumption superposition value Zdj, and then calculating the conflict degree index Zct of each time window by comparing it with the load threshold Fz of the energy supply system, it is possible to quickly and clearly determine whether there is a superimposed energy consumption conflict in the current time window. Its core function is that it can not only achieve a quantitative assessment of energy consumption conflict at the time window level, but also provide a clear logical basis for whether to enter the optimization control process. Compared with traditional manual investigation and experience-based peak shifting strategies, the conflict degree index Zct realizes a digital expression of the concentrated load trend of complex processes, has the ability of predictive judgment and closed-loop triggering, and is a key central variable supporting the dynamic control and adaptive optimization of enterprise energy.

[0039] In this embodiment, by introducing the construction and calculation of the conflict degree index Zct, the matching relationship between the superposition strength of energy consumption coupling and energy supply capacity is quantitatively determined, improving the enterprise's ability to predict overload risks in complex parallel process environments. By associating the superposition value of coupled energy consumption Zdj, which is composed of all process nodes in each time window, with the maximum power Fz that the enterprise's production line can stably output in the same time window, and constructing the conflict degree index Zct after dimensionless processing, the energy consumption conflict risk of each time window can be dynamically quantified on a unified scale. The significant advantage of this mechanism is that it no longer relies solely on the high or low load of a single device, but is based on the sum of the behaviors of all coupled pairs under the graph-based energy consumption interaction model, comprehensively reflecting whether the overall process system faces a mismatch between energy supply and demand in a certain time window. Especially in manufacturing scenarios with frequent multi-node coupling, highly concentrated load peaks, and tight task scheduling cycles, this step can achieve real-time identification and early warning of energy consumption conflicts, effectively avoiding the scheduling rigidity and resource waste caused by previous methods such as experience-based peak shifting and static power limiting, and providing a scientific basis for stable production line operation and energy scheduling strategy adjustment.

[0040] Example 5 Please refer to Figure 1 Specifically, the S4 steps include: S41. Based on the conflict degree index Zct values ​​obtained in step S31 for each time window, determine whether there is an overload risk due to the superposition of energy consumption conflicts during the operation of multiple process nodes within the corresponding time window, and execute the corresponding energy optimization strategy. The specific content is as follows: If the conflict index Zct of the corresponding time window is greater than 1, it means that the maximum power that the energy supply system can stably output within the current time window cannot meet the operation of multiple process nodes. In this case, it is judged that there is an overload risk in the operation of multiple process nodes within the corresponding time window. At this time, the time window corresponding to the conflict index Zct is recorded as the overload time window, and an energy optimization command is issued to the outside. If the conflict index Zct of the corresponding time window is less than or equal to 1, it means that the maximum power that the energy supply system can stably output within the current time window can meet the operation of multiple process nodes. Therefore, it is determined that there is no risk of overload in the operation of multiple process nodes within the corresponding time window. The time window corresponding to the conflict index Zct is recorded as the normal time window set. When all time windows are marked as normal time windows, a normal operation signal is sent out. S42. Upon receiving the energy optimization instruction, based on the process behavior data map... As input data for the energy consumption coupling analysis model, combined with heuristic search algorithm, local disturbance rearrangement mechanism and conflict sensitivity weight adjustment method, the concurrent start-up count of the corresponding process node in the overload time window, the real-time high power trigger point and the process chain synchronization lag time are updated and adjusted, which are used as new input data for the energy consumption coupling analysis model. Steps S2 and S3 are repeated until a normal operation signal is issued.

[0041] It should be noted that when multiple process nodes are found to have overlapping energy consumption conflicts within a certain time window, causing the energy system load threshold to be exceeded, a static strategy is no longer adopted. Instead, an energy consumption coupling analysis model based on the process behavior data map XW is used to intelligently and dynamically adjust the key behavioral parameters of the conflicting nodes, achieving precise rescheduling of production cycles. Specifically, this optimization process is completed collaboratively by three types of algorithm mechanisms: First, a heuristic search algorithm is mainly used to quickly locate feasible initial solutions for adjustment schemes in a complex parameter space, prioritizing the identification of nodes that can postpone startup without affecting downstream production lines in the concurrent startup count Cqd; second, a local perturbation rearrangement mechanism is used to further refine the initial solution based on partial... The startup time, peak power trigger point Ngg, or synchronization lag time Tzh of the nodes are slightly perturbed to simulate various permutations and combinations, and the changing trend of energy consumption coupling is observed. Finally, the conflict sensitivity weight adjustment method assigns different weights to the degree of conflict impact of different nodes. For example, stronger control priority is applied to nodes in bottleneck sections or those with frequent high power superposition, making the overall adjustment more targeted. After optimization by the above algorithm combination, it is used as new input data for the energy consumption coupling analysis model, and the calculation process of steps S2 and S3 is re-executed. Finally, it is determined whether the conflict degree index Zct of all time windows has dropped to an acceptable range, and a normal operation signal is issued to achieve closed-loop dynamic optimization and adjustment of energy consumption conflict.

[0042] In this embodiment, by introducing a dynamic discrimination mechanism based on the conflict degree index Zct and a responsive energy optimization strategy, closed-loop intelligent control and adaptive adjustment of energy consumption conflict scenarios are achieved, significantly improving the energy use safety and operational flexibility of the production line under multi-process coupling conditions. This step no longer relies on the traditional experience-based energy consumption adjustment mode, but instead uses the dynamic numerical judgment of the conflict degree index Zct. When the conflict degree index Zct > 1 in the corresponding time window, it can accurately identify the overload time window with the risk of superimposed energy consumption overload and trigger energy optimization commands in real time. Compared with the static load limit mechanism, it can complete predictive judgment before abnormal process load interaction behavior occurs, thereby implementing pre-intervention in the scheduling stage and effectively avoiding local tripping, equipment abnormalities, or production line disruptions caused by energy consumption peak coupling. The system mitigates the significant risk of system interruption. Upon receiving optimization instructions, based on a graph-based process behavior model, and combining heuristic search algorithms, local perturbation rearrangement mechanisms, and conflict sensitivity weight adjustment methods, it dynamically adjusts key input parameters such as the number of concurrent startups (Cqd), the real-time high-power trigger point (Ngg), and the process chain synchronization lag time (Tzh). This adjustment mechanism not only possesses the capability for refined intervention but also directly acts on the scheduling input at the model level, achieving precise coupling between energy optimization control logic and process behavior structure. This approach significantly differs from traditional optimization methods that only constrain outputs at the result level, realizing a vertical closed loop from behavior recognition to parameter intervention, ensuring that the adjustment strategy truly aligns with the actual process characteristics. The system iteratively reconstructs steps S2 and S3 until a normal operation signal is issued.

[0043] Example 6 Please refer to Figure 1 and Figure 2 Specifically: an enterprise energy intelligent optimization system based on big data analysis, including a graph construction module, an overlay analysis module, a conflict analysis module, and an energy optimization module; The graph construction module uses the process behavior data graph XW to represent the energy consumption behavior between various process nodes during the operation of multiple process nodes within the enterprise's precision electronics manufacturing production line, and uniformly marks each process node in the same time window in the graph. The superposition analysis module monitors the energy consumption level of each process node within each time window. Any two process nodes within the same time window are recorded as a pair, and the superposition effect of the coupling energy consumption of all pairs within each time window is analyzed. The conflict analysis module analyzes the degree of conflict of energy consumption superposition during the operation of multiple process nodes in each time window based on the maximum power that the energy supply system of the enterprise production line can stably output within a unit time window, and obtains the conflict degree index Zct for each time window. The energy optimization module determines whether there is an overload risk due to the superposition of energy consumption conflicts during the operation of multiple process nodes within the corresponding time window based on the value of the conflict degree index Zct for each time window, and then executes the corresponding energy optimization strategy.

[0044] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A smart energy optimization method for enterprises based on big data analysis, characterized by: Includes the following steps: S1. During the operation of multiple process nodes within the precision electronics manufacturing production line of an enterprise, the energy consumption behavior between each process node is expressed in a graph structure using the process behavior data map XW, and each process node in the same time window in the map is uniformly marked. S2. Monitor the energy consumption level of each process node within each time window, and take any two process nodes within the same time window as a pair, and analyze the superposition effect of the coupling energy consumption of all pairs within each time window. S3. Based on the maximum power that the energy supply system of the enterprise's production line can stably output within a unit time window, analyze the degree of conflict of energy consumption superposition during the operation of multiple process nodes within each time window, and obtain the conflict degree index Zct for each time window. S4. Based on the conflict degree index Zct value of each time window, determine whether there is an overload risk of energy consumption superposition conflict during the operation of multiple process nodes within the corresponding time window, and execute the corresponding energy optimization strategy.

2. The enterprise energy intelligent optimization method based on big data analysis according to claim 1, characterized in that: The specific steps in S1 include: S11. During the operation of multiple process nodes within the enterprise's precision electronics manufacturing production line, based on IoT architecture, edge computing, and high-frequency sensing devices, the concurrent start count, real-time high-power trigger point, and process chain synchronization lag time of each process node within each time window are obtained. The specific process includes: S111. By using the event trigger logs and PLC instruction streams embedded in the equipment control system, identify the concurrent startup behavior of multiple process nodes within a unit time window, perform real-time statistics on the trigger frequency of equipment state changes within a unit time window, form statistics and record them, and obtain the concurrent startup count Cqd of each process node within each time window. S112. Based on the high-frequency energy consumption metering device integrated in the process control system, the energy consumption of each process node is sampled in real time, and the energy consumption derivative threshold within the sliding window is set. The trigger point of the power change in the sampled signal within the unit time window is identified. The edge computing unit completes the preliminary judgment and uploads it to the cloud for processing to obtain the real-time high power trigger point Ngg of each process node within each time window. S113. Based on the start and finish timestamps of each process, the interval between task delivery and task receipt is automatically recorded each time a product is transferred from the upstream process node to the downstream process node. The average delay is extracted based on statistical rules to form the process chain synchronization lag time Tzh of each process node within each time window.

3. The enterprise energy intelligent optimization method based on big data analysis according to claim 2, characterized in that: The specific steps in S1 also include: S12. Based on the concurrent start-up count Cqd of each process node in each time window, the real-time high-power trigger point Ngg, and the synchronization lag time Tzh of the process chain, a sliding window matching algorithm is introduced to reorganize the timing and form a process behavior data map XW that scrolls in a sliding window manner, which is used as the input data of the energy consumption coupling analysis model. S13. Based on the constructed process behavior data map XW, the energy consumption behavior between each process node is expressed in a graph structure, and each process node in the same time window in the graph is uniformly marked.

4. The enterprise energy intelligent optimization method based on big data analysis according to claim 3, characterized in that: The specific steps in S2 include: S21. After uniformly marking each process node based on the same time window, the energy consumption level of each process node in each time window is monitored by the high-frequency energy consumption metering device integrated in the process control system, and the energy consumption value Enh of each process node in each time window is obtained. The high-frequency energy consumption metering device includes a three-phase smart energy meter, a power quality analyzer, a heat mass flow meter and an edge acquisition device.

5. The enterprise energy intelligent optimization method based on big data analysis according to claim 4, characterized in that: S2 also includes the following specific steps: S22. Let any two process nodes within the same time window be denoted as a pair (i, j). Based on the energy consumption value Enh of each process node within each time window obtained in step S11, and combined with the time series similarity measurement algorithm, analyze the degree of resource usage overlap between any two process nodes within the same time window, and obtain the resource usage overlap coefficient Xcd of each pair within each time window. The specific formula is as follows: ； In the formula, Let be the resource usage overlap coefficient of the combination pair of process node i and process node j within the k-th time window, where It represents the minimum value of the intersection of the actual energy consumption of the i-th process node and the j-th process node within the k-th time window. It is represented as the maximum value of the union of energy consumption of the i-th process node and the j-th process node within the k-th time window.

6. The enterprise energy intelligent optimization method based on big data analysis according to claim 5, characterized in that: S2 also includes the following specific steps: S23. Based on the timestamp recording function in the process control system, and combined with the resource usage overlap coefficient Xcd of each pair in each time window obtained in step S22, record the operation end timestamp and operation start timestamp of the two process nodes in the corresponding pair in the corresponding time window, calculate the operation time difference of the two process nodes in each pair, and record it as the execution operation time difference Tsc of each pair in each time window. S24. Correlate the resource usage overlap coefficient Xcd of each pair within each time window with the corresponding execution operation time difference Tsc. After dimensionless processing, analyze the energy consumption coupling degree of any two process nodes executing operations within the same time window, and obtain the execution energy consumption coupling coefficient Xoh of each pair within each time window. Specifically, it is obtained through the following formula: ； In the formula, Let be the execution energy coupling coefficient of the pair of process nodes i and j within the k-th time window, where This is expressed as the operation time difference adjustment factor. It is represented as an exponential function.

7. The enterprise energy intelligent optimization method based on big data analysis according to claim 6, characterized in that: S2 also includes the following specific steps: S25. Correlate the execution energy consumption coupling coefficient Xoh of each pair within each time window obtained in step S24 with the energy consumption value Enh of each process node within each time window obtained in step S11, analyze the coupling energy consumption superposition effect of all pairs within each time window, and obtain the coupling energy consumption superposition value Zdj of each time window, which is specifically obtained through the following formula: ； In the formula, Let be the sum of the coupled energy consumption for the k-th time window, where This represents the number of pairings between any two process nodes within the same time window.

8. The enterprise energy intelligent optimization method based on big data analysis according to claim 7, characterized in that: The specific steps in S3 include: S31. The maximum power that the energy supply system of the enterprise's production line can stably output within a unit time window is recorded as the load threshold Fz. It is correlated with the coupled energy consumption superposition value Zdj of each time window obtained in step S25. After dimensionless processing, the degree of conflict of energy consumption superposition during the operation of multiple process nodes within each time window is analyzed to obtain the conflict degree index Zct of each time window, which is obtained by the following formula: ； In the formula, It is represented as the conflict level index for the k-th time window.

9. The enterprise energy intelligent optimization method based on big data analysis according to claim 8, characterized in that: The specific steps of S4 include: S41. Based on the conflict degree index Zct values ​​obtained in step S31 for each time window, determine whether there is an overload risk due to the superposition of energy consumption conflicts during the operation of multiple process nodes within the corresponding time window, and execute the corresponding energy optimization strategy. The specific content is as follows: If the conflict index Zct of the corresponding time window is greater than 1, it means that the maximum power that the energy supply system can stably output within the current time window cannot meet the operation of multiple process nodes. In this case, it is judged that there is an overload risk in the operation of multiple process nodes within the corresponding time window. At this time, the time window corresponding to the conflict index Zct is recorded as the overload time window, and an energy optimization command is issued to the outside. If the conflict index Zct of the corresponding time window is less than or equal to 1, it means that the maximum power that the energy supply system can stably output within the current time window can meet the operation of multiple process nodes. Therefore, it is determined that there is no risk of overload in the operation of multiple process nodes within the corresponding time window. The time window corresponding to the conflict index Zct is recorded as the normal time window set. When all time windows are marked as normal time windows, a normal operation signal is sent out. S42. Upon receiving the energy optimization instruction, based on the process behavior data map... As input data for the energy consumption coupling analysis model, combined with heuristic search algorithm, local disturbance rearrangement mechanism and conflict sensitivity weight adjustment method, the concurrent start-up count of the corresponding process node in the overload time window, the real-time high power trigger point and the process chain synchronization lag time are updated and adjusted, which are used as new input data for the energy consumption coupling analysis model. Steps S2 and S3 are repeated until a normal operation signal is issued.

10. An enterprise energy intelligent optimization system based on big data analysis, used to implement the enterprise energy intelligent optimization method based on big data analysis as described in any one of claims 1 to 9, characterized in that: It includes a map construction module, an overlay analysis module, a conflict analysis module, and an energy optimization module; The graph construction module uses the process behavior data graph XW to represent the energy consumption behavior between various process nodes during the operation of multiple process nodes within the enterprise's precision electronics manufacturing production line, and uniformly marks each process node in the same time window in the graph. The superposition analysis module monitors the energy consumption level of each process node within each time window. Any two process nodes within the same time window are recorded as a pair, and the superposition effect of the coupling energy consumption of all pairs within each time window is analyzed. The conflict analysis module analyzes the degree of conflict of energy consumption superposition during the operation of multiple process nodes in each time window based on the maximum power that the energy supply system of the enterprise production line can stably output within a unit time window, and obtains the conflict degree index Zct for each time window. The energy optimization module determines whether there is an overload risk due to the superposition of energy consumption conflicts during the operation of multiple process nodes within the corresponding time window based on the value of the conflict degree index Zct for each time window, and then executes the corresponding energy optimization strategy.

Images