Multi-source data fused power grid intelligent supply chain operation efficiency evaluation and intelligent decision-making system

Abstract

The invention discloses a power grid intelligent supply chain operation efficiency evaluation and intelligent decision-making system fusing multi-source data, and relates to the technical field of power grid supply chain management, in particular to the power grid intelligent supply chain operation efficiency evaluation and intelligent decision-making system fusing the multi-source data. Comprising a heterogeneous data access gateway, a data quality governance engine, a dynamic semantic association module, a supply chain operation efficiency multi-dimensional evaluation model, an intelligent decision generation module and a closed-loop feedback and self-learning optimization module. The system uniformly accesses and manages multi-source heterogeneous data, constructs a dynamic association data network, carries out efficiency evaluation and deviation diagnosis based on a multi-dimensional index system and a dynamic datum line, generates decision suggestions in combination with a rule engine and multi-objective optimization simulation, supports hierarchical autonomous execution, and continuously feeds back optimization evaluation weights and decision rules, so that the decision suggestions of the multi-dimensional index system and the dynamic datum line are optimized. Real-time sensing, intelligent diagnosis, predictive decision making and self-adaptive optimization of the operation state of the supply chain are achieved, and the operation efficiency, toughness and decision making scientificity of the supply chain are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of power grid supply chain management technology, specifically to a smart power grid supply chain operation efficiency assessment and intelligent decision-making system that integrates multi-source data. Background Technology

[0002] As the lifeline and critical infrastructure of national energy supply, the safe, stable, and efficient operation of the power grid is of paramount importance. The power grid supply chain, as the core support system for power grid construction, operation, maintenance, and emergency repair, involves a wide variety of materials, a vast supplier network, extensive warehousing nodes, and complex logistics routes, and is closely coupled with core power grid operations such as production, infrastructure, and operation. With the accelerated construction of new power systems and the increasing demands for lean management of power grid assets under the "dual carbon" target, traditional power grid supply chain management models are facing unprecedented challenges.

[0003] First, at the data level, the power grid supply chain has long suffered from severe "data silos" and "data heterogeneity" problems. Massive amounts of data, including master data on materials, supplier information, project demand plans, warehouse inventory status, in-transit logistics information, and equipment sensor readings, are scattered across different independent systems. These data formats vary widely, standards differ, and there is a lack of effective integration and governance mechanisms. This makes it difficult for management to obtain a comprehensive, real-time, and unified view of the supply chain. Decisions are often based on partial, lagging, or even inconsistent data, severely impacting the scientific rigor and accuracy of decision-making.

[0004] Secondly, at the operational assessment level, traditional supply chain efficiency assessments often rely on static, isolated financial or operational indicators, lacking a multi-dimensional, quantifiable, and comprehensive assessment system that can fully reflect supply chain resource efficiency, process agility, collaborative response capabilities, and risk resilience. The assessment process is often lagging, unable to provide dynamic early warnings and trend judgments for operational anomalies, and thus failing to support pre-emptive prevention and in-process control.

[0005] Furthermore, at the decision support level, current management decisions heavily rely on the personal experience of managers, lacking systematic intelligent analysis and support. When inventory backlogs, delivery delays, or supplier risks occur, manual analysis of numerous reports is typically required to pinpoint the root cause of the problem and formulate response strategies based on experience. This approach is slow to respond and struggles to comprehensively balance multiple objectives such as cost, time, service, and risk. It also fails to quantitatively predict or simulate the future effects of decision-making solutions, representing a typical "experience-driven" rather than "data and model-driven" decision-making model.

[0006] Finally, regarding system evolution capabilities, most existing information systems lack self-learning and adaptive optimization mechanisms. The effectiveness of decisions and whether evaluation models align with actual business changes require human intervention and periodic adjustments. The system itself cannot learn from the successes or failures of historical decisions, nor can it achieve continuous iterative optimization of evaluation models and decision-making strategies, thus limiting the system's long-term applicability and intelligence level. Summary of the Invention

[0007] The purpose of this invention is to provide a smart power grid supply chain operation efficiency assessment and intelligent decision-making system that integrates multi-source data. By integrating and dynamically correlating multi-source data, it breaks down information silos; through multi-dimensional dynamic assessment and precise deviation diagnosis, it achieves real-time insight and early warning of supply chain status; through intelligent decision generation and multi-objective simulation and deduction, it provides scientific and forward-looking decision support; and finally, through a closed-loop self-learning mechanism, the system has the ability to continuously optimize, significantly improving the operational efficiency, resilience, and intelligence level of the power grid supply chain.

[0008] To achieve the above objectives, this invention employs the following technical solution: a smart power grid supply chain operation efficiency assessment and intelligent decision-making system integrating multi-source data. The core function of the heterogeneous data access gateway lies in its powerful data fusion capability. It is not a simple data pipeline, but an intelligent access layer with complex protocol conversion and semantic parsing capabilities. Faced with structured tabular data from the power grid material master data management platform, real-time time-series data generated by IoT device sensing networks, semi-structured location messages from logistics tracking systems, and heterogeneous data streams from external macroeconomic and social data platforms, the gateway's built-in standardized adapter interface library can automatically identify the data source format and perform preliminary decoding and normalization. For example, for sensors from different brands, even if their data reporting protocols differ, the adapter can uniformly convert them into standardized data objects within the system and add a unified timestamp and data source identifier. This design ensures that fragmented data from the physical world and different information systems can be smoothly and unambiguously accessed into the system, laying a solid data foundation for subsequent in-depth analysis and solving the data silos and format barriers problems at the forefront of smart supply chain construction.

[0009] Furthermore, the innovation of the multi-dimensional assessment model for supply chain operational efficiency lies in its construction of a comprehensive and balanced assessment framework. This model deeply understands that supply chain efficiency cannot be summarized by a single indicator. Therefore, it comprehensively considers four core dimensions: resource efficiency, process agility, collaborative response, and risk resilience. The resource efficiency dimension focuses on asset utilization levels, such as revealing capital flow efficiency through inventory turnover rate. The process agility dimension measures the speed and stability of supply chain execution, such as order fulfillment cycle directly related to customer satisfaction. The collaborative response dimension assesses internal and external collaboration efficiency, such as demand forecast accuracy affecting the planning accuracy of the entire chain. The risk resilience dimension proactively assesses the system's ability to withstand interference, such as supplier concentration analysis aimed at avoiding supply chain disruption risks. Crucially, the adaptive weight allocation algorithm used in the model can dynamically adjust the relative importance of each dimension indicator based on factors such as peak and off-peak seasons and strategic priorities (such as cost priority or delivery priority). This makes the final comprehensive efficiency index and radar chart more realistic and accurate in reflecting operational health under specific scenarios, guiding management activities to focus on the main contradictions.

[0010] Furthermore, the efficiency deviation calculation method based on dynamic baselines upgrades efficiency assessment from a static snapshot to dynamic tracking. The core of this method lies in acknowledging that the standard of "excellence" is not static but changes with the internal and external environment. The dynamic baseline it establishes not only references the company's own historical best practices and average levels but also incorporates industry benchmarks and expected changes brought about by real-time market demand fluctuations. The deviation analysis performed by the calculation module goes beyond simple numerical comparisons and focuses more on trend judgment. For example, for a decline in inventory turnover, the system will not only report the deviation value but also analyze whether the decline is a short-term fluctuation or a continuous deterioration trend, and comprehensively judge its severity level through fuzzy logic. At the same time, this method has a powerful root cause tracing capability, which can locate the deviation to a specific warehousing node, material category, or supplier, thereby directly transforming the assessment results into actionable insights, guiding managers to quickly identify the source of the problem, and realizing the leap from "discovering the problem" to "locating the problem".

[0011] Furthermore, the core value of the supply chain intelligent decision generation module lies in transforming evaluation results into actionable solutions. The decision tree embedded in this module integrates the clear logic and constraint satisfaction theory of the business rule engine to handle complex trade-offs. When receiving performance deviation diagnosis results, the decision tree generates a set of feasible and mutually exclusive decision options based on hundreds of preset business rules and combined with the real-time constraints in the current system. Each option is accompanied by detailed execution steps and expected key performance indicator improvement values. For example, for the risk of delivery delays, the decision tree may simultaneously provide multiple solutions such as "urge the current carrier", "activate the alternative logistics provider", and "initiate transfer from a nearby warehouse", clearly listing the cost increment, expected time savings, and potential risks of each solution. This provides managers with a structured decision support menu, significantly improving the scientific nature and efficiency of decision-making.

[0012] Furthermore, the introduction of a multi-objective optimization simulator greatly enhances the foresight and scientific rigor of decision-making. This simulator treats the supply chain as a complex system, acknowledging that it is difficult for any decision to simultaneously achieve optimal results across multiple objectives such as cost, time, service level, and risk control; trade-offs are inevitable. When making decision recommendations, the simulator allows managers to set priority weights for each objective based on current strategic priorities. Subsequently, it uses historical data and the current state to construct a simplified supply chain simulation environment, performing thousands of scenario simulations on candidate decision-making options. These simulations consider uncertainties such as demand fluctuations, transportation delays, and price changes, ultimately quantifying and predicting the comprehensive impact of each option on various objectives under different possible futures. The results are presented in intuitive chart forms, such as parallel coordinate graphs or trade-off curves, enabling managers to clearly see the gains and losses behind different choices, thereby making more forward-looking decisions that align with long-term interests, rather than passively reacting based solely on past experience.

[0013] Furthermore, the closed-loop feedback and self-learning optimization module is the "brain" that enables the system to continuously evolve. By establishing a complete data loop from decision generation and execution to effect evaluation, this module enables the system to improve itself. Once a decision suggestion is adopted and implemented, the module continuously tracks the actual changes in relevant efficiency indicators and compares the actual effect with the prediction at the time the decision was generated. The built-in causal analysis model uses statistical methods to try to eliminate other interfering factors and assess the strength of the causal relationship between the decision action and the indicator changes. Based on this feedback data, the module automatically fine-tunes the indicator weight allocation in the evaluation model or optimizes the rule threshold parameters in the decision tree. For example, if historical data shows that the effect of a certain optimization strategy is consistently lower than expected in practice, the system will correspondingly reduce the recommendation priority of such strategies. This mechanism enables the system to adapt to changes in the business environment, continuously improve the accuracy of evaluation and decision-making, and form a virtuous cycle of becoming smarter the more it is used.

[0014] Furthermore, the edge computing capabilities of heterogeneous data access gateways are an effective strategy for dealing with the deluge of IoT data. In the power grid supply chain, especially in warehousing and transportation, a large number of sensors are deployed to monitor information such as temperature, humidity, vibration, and location in real time. The data volume is huge and the value density is low. The gateway performs preprocessing near the data source, such as threshold judgment on temperature sensor data, and only reports abnormal events when the data exceeds the range or fluctuates drastically. It also aggregates GPS data and reports track points periodically instead of continuously streaming it. This edge computing model brings multiple benefits: First, it greatly reduces the computing and storage pressure on the central server and reduces IT infrastructure costs. Second, it reduces network bandwidth consumption, which is particularly suitable for remote warehouses or mobile transportation environments with poor network conditions. Finally, and most importantly, it enables local real-time response. In case of emergencies such as fire alarms or temperature control failures, local alarms and emergency mechanisms can be triggered within milliseconds, significantly improving the real-time risk management capabilities of the supply chain.

[0015] Furthermore, the visualization through an interactive digital twin cockpit greatly enhances the intuitiveness of the assessment results and the management's penetrating power. This cockpit is not simply a collection of two-dimensional charts, but rather a three-dimensional virtual model that is highly mirrored with the physical supply chain. In this model, managers can have a panoramic view of the entire supply chain network. Various warehouses, transportation routes, inventory levels, and order statuses are dynamically presented as visual elements. Users can directly click on virtual warehouses to view detailed indicators such as space utilization and inventory turnover. They can click on a transportation route to track the location of materials and estimated arrival time in real time. They can also slide along the timeline to replay the evolution of historical operational status. This immersive interactive experience closely links abstract assessment indices with specific physical entities and business processes, enabling managers to quickly understand the business meaning behind complex data and achieve seamless drilling from macro to micro levels, greatly enhancing the intuitiveness of decision-making and the ability to perceive context.

[0016] Furthermore, the hierarchical autonomous execution capability of the intelligent decision generation module is a key step in realizing supply chain intelligence. This function elevates decision support to a new level of decision automation. The system establishes a clear decision authority matrix based on the risk level, complexity, and standardization of decision-making matters. For high-frequency routine operations with clear rules and controllable risks, such as routine replenishment based on the safety stock model and supplier payments that meet preset conditions, the system can directly convert decision suggestions into API instructions after a built-in electronic approval process, and issue them to the procurement system or warehouse management system for execution. This "unmanned" operation not only frees managers from a large number of repetitive and transactional tasks, but also significantly improves the response speed and execution efficiency of the supply chain, reduces human operation delays and errors, and ensures the auditability of all automated decisions. At the same time, the system will fully record the entire process of all automated decisions, ensuring the auditability of operations, allowing managers to focus on handling more complex anomalies and strategic decisions.

[0017] Furthermore, the dynamic knowledge base functions to transform discrete operational experiences into sustainable organizational assets. Driven by a closed-loop feedback module, this knowledge base continuously accumulates and structurally stores various types of knowledge from the operational process. It not only records successful decision-making cases and their effect data, but more importantly, it systematically includes various abnormal patterns of performance deviations, their root causes after in-depth analysis, and effective coping strategies proven in practice. This knowledge is indexed and linked through an ontology model, forming a valuable "corporate memory." When the system detects an abnormal pattern similar to historical cases again, it can intelligently retrieve relevant coping strategies and historical effect records from the knowledge base as strong evidence for decision-making recommendations. This not only accelerates the problem-solving process, but more importantly, it enables individual lessons learned to be shared and reused at the organizational level, effectively preventing the recurrence of similar problems, thereby continuously improving the collective learning ability and decision-making resilience of the entire supply chain organization.

[0018] This invention provides a smart power grid supply chain operation efficiency assessment and intelligent decision-making system that integrates multi-source data, which has the following beneficial effects: 1. Data level: It has achieved a leap from "heterogeneous silos" to "unified high-quality data base", laying the foundation for accurate decision-making.

[0019] Enhancing the breadth and efficiency of data integration: Through the standardized interface of the heterogeneous data access gateway, the system can seamlessly access heterogeneous data from multiple sources inside and outside, breaking the "information silos" situation of scattered data sources and different formats in the traditional supply chain, realizing the comprehensive aggregation of data resources, and providing the possibility for analysis and decision-making from a global perspective.

[0020] Ensuring data quality and credibility: The data quality governance engine ensures the integrity, consistency, and accuracy of incoming data through automated cleaning, repair, and labeling, generating quantifiable data health reports. This fundamentally avoids the problem of "garbage in, garbage out," greatly improving the reliability and credibility of subsequent assessments and decision-making results.

[0021] Enhancing Data Intelligence and Relevance: The dynamic semantic association module, based on a domain ontology model, transforms raw data into a dynamic relational network with rich semantic relationships. This enables the system to understand the business meaning behind the data, rather than simply processing numbers, laying the foundation for in-depth analysis and intelligent insights.

[0022] At the assessment level, it has achieved an upgrade from "static reports" to "dynamic and proactive diagnosis", enhancing the ability to perceive the status.

[0023] Achieving multidimensionality and comprehensiveness in assessment: The constructed assessment system, which considers four dimensions—resource efficiency, process agility, collaborative response, and risk resilience—can comprehensively and three-dimensionally reflect the operational status of the supply chain, avoiding the one-sidedness of assessment based on a single indicator.

[0024] Introducing dynamic and contextualized perception: The performance deviation calculation method based on dynamic baselines transforms evaluation from a simple comparison with fixed targets to contextualized analysis that incorporates historical data, industry benchmarks, and real-time market fluctuations. The system can identify trends and determine severity levels, thereby shifting from "post-event statistics" to "in-event early warning" and "pre-event prediction," significantly enhancing management initiative.

[0025] Enhancing the interpretability and operability of evaluation results: The evaluation model can pinpoint the key data sources and business processes that cause deviations, making efficiency issues not only quantifiable but also traceable, providing a clear entry point and direction for subsequent decision optimization.

[0026] At the decision-making level: it has achieved a leap from "experience-driven" to "data and model-driven", optimizing the quality and efficiency of decision-making.

[0027] Improving the scientific rigor and automation of decision-making: The intelligent decision generation module, based on a rule engine and constraint satisfaction theory, can automatically generate multiple priority-ranked optimization solutions for specific performance deviation problems and predict their impact. This reduces over-reliance on managers' personal experience, making the decision-making process more scientific and standardized.

[0028] Supporting multi-objective trade-offs and predictive decision-making: The introduction of a multi-objective optimization simulator is a major breakthrough. It allows managers to weigh multiple objectives such as cost, time, and service, and to predict the combined effects of different decisions under various future market scenarios through simulation and deduction. This achieves a qualitative leap from "reactive decision-making based on historical data" to "predictive decision-making based on future scenarios," significantly reducing decision-making risks.

[0029] Enhancing operational efficiency and execution agility: Deep integration of the system and execution system, along with hierarchical autonomous execution capabilities, enables the automatic execution of high-frequency decisions with clearly defined rules, achieving "unmanned" operation. This not only frees managers from tedious tasks, allowing them to focus on strategic work, but also significantly shortens the decision-making-execution cycle and improves the supply chain's responsiveness to market changes.

[0030] At the system evolution level: it has achieved the evolution from a "fixed tool" to a "self-learning, self-optimizing organism", ensuring long-term value.

[0031] A closed loop of continuous optimization is formed: The closed-loop feedback and self-learning optimization module compares the expected decision with the actual effect, and uses the causal analysis model to continuously fine-tune the evaluation weights and decision rules, so that the system can learn from actual operation and become "smarter" the more it is used, and the accuracy of evaluation and decision-making will continue to improve over time.

[0032] Facilitating the accumulation and reuse of organizational knowledge: The dynamic knowledge base transforms successful decision-making cases, abnormal patterns, and effective strategies into reusable structured knowledge. When similar problems recur, the system can prioritize recommending historically validated solutions, transforming individual experience into organizational wisdom. This effectively enhances the collective decision-making capabilities and resilience of the entire supply chain system, preventing the recurrence of mistakes.

[0033] In terms of user experience and system performance: intuitive interaction and efficient operation have been achieved.

[0034] Enhance the intuitiveness and insight of management: Through the three-dimensional visualization of the digital twin cockpit, managers can intuitively and immersively grasp the overall picture of the supply chain, and through drill-down operations, they can go directly from macro indicators to micro data, which greatly improves the intuitiveness and insight of management decisions.

[0035] Optimize system resources and real-time performance: The edge computing capabilities of the heterogeneous data access gateway enable preprocessing at the data source, effectively reducing the load on the central server and network bandwidth pressure, ensuring the system's ability to process high real-time data such as IoT data, and ensuring the efficient and stable operation of the entire system. Attached Figure Description

[0036] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings in the following description are merely exemplary, and those skilled in the art can derive other embodiments based on the provided drawings without creative effort.

[0037] Figure 1 This is a flowchart illustrating the data integration and governance process of this invention. Figure 2 This is a flowchart of the efficiency evaluation process for this invention; Figure 3 This is a flowchart of the intelligent decision generation process of the present invention; Figure 4 This is a flowchart illustrating the closed-loop feedback and self-learning process of this invention. Figure 5 This is a flowchart illustrating the visualization and autonomous execution process of this invention. Detailed Implementation

[0038] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numerals in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this disclosure. Rather, they are merely examples of apparatuses consistent with some aspects of this disclosure as detailed in the appended claims.

[0039] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0040] Example 1: Addressing scenarios with low inventory turnover rates for specific materials Scenario Description: In a regional warehouse of a power grid company, the inventory turnover rate of special cables used for distribution network upgrades has been lower than the dynamic baseline set by the system for two consecutive weeks, indicating a risk of overstocking.

[0041] System application process: Data Access and Evaluation: The heterogeneous data access gateway collects real-time inventory data and IoT-sensed inbound and outbound records for the warehouse. After calculation by the multi-dimensional supply chain operation efficiency evaluation model, an early warning is triggered in the resource efficiency dimension, indicating a significant deviation in the inventory turnover rate of this category of materials.

[0042] Intelligent diagnostics: Analysis using the dynamic baseline calculation method indicates that the deviation is related to a recent delay in the construction of a key project, leading to a lag in demand. The system identifies the associated project requirements and supplier information.

[0043] Decision Generation and Execution: The intelligent decision generation module automatically generates two priority options based on the rules engine: First, initiate a collaborative inventory replenishment agreement with the supplier to postpone subsequent batches of goods; second, prompt managers in the supply chain digital twin dashboard to check for available allocation needs in nearby areas. After managers select the first option, the system sends an instruction to the supplier relationship management system to adjust the supply plan through a deeply integrated execution interface.

[0044] Application results: It prevented further inventory backlog and reduced capital occupation costs. The closed-loop feedback module confirmed that the inventory turnover rate returned to the normal range in subsequent cycles and recorded this decision scenario and its effects in the dynamic knowledge base.

[0045] Example 2: Scenario for improving the on-time delivery rate of materials for major projects Scenario Description: A new substation project has entered a critical stage, with extremely high requirements for the timely delivery of the main transformer. At this time, the logistics tracking system indicates that weather delays may occur during transportation.

[0046] System application process: Data Access and Evaluation: Real-time location data from the logistics tracking system and external meteorological data were integrated. The on-time delivery rate, a key performance indicator for process agility, began to show fluctuations and warnings.

[0047] Simulation and simulation: Managers launched a multi-objective optimization simulator, setting "time" as the highest priority objective. Based on data such as current transportation location, alternative routes, and alternative supplier capacity, the simulator simulated various scenarios: the risk of a 3-day delay on the original route, the increased cost of activating the alternative route but the possibility of on-time arrival, and the activation of the emergency supply plan from the alternative supplier, etc., and quantified the impact of each scenario on the project schedule.

[0048] Decision Support: Based on the simulation results, the system prioritized recommending the option of "activating alternative routes and strengthening collaboration with logistics providers," and estimated the probability that this option could guarantee on-time delivery. Managers made their decisions based on this clear and predictive analysis.

[0049] Application Results: Through proactive intervention, materials were delivered to the construction site on time, ensuring the smooth progress of key project milestones. The decision-making logic and external event data from this successful response were incorporated into a dynamic knowledge base for future scenario matching for similar risks.

[0050] Example 3: Optimizing Supplier Collaborative Response to Address Sudden Demand Scenario Description: A sudden natural disaster has led to a surge in demand for emergency relief supplies, requiring the rapid coordination of production capacity and distribution resources from multiple suppliers.

[0051] System application process: Data Association: The dynamic semantic association module quickly associates and maps the demand instructions from the emergency command platform with the list of qualified suppliers in the supplier relationship management system, their current capacity load, and the spot inventory in the warehouse network.

[0052] Multi-dimensional assessment: The risk resilience and collaborative response dimensions of the assessment model are analyzed simultaneously. The system not only evaluates the responsiveness of individual suppliers, but also comprehensively calculates the concentration risk of the supplier group and identifies weak links in key supply paths.

[0053] Optimization Solution Generation: Based on constraint satisfaction theory (such as capacity limits and transportation capacity), the decision-making module generates a combined solution: Supplier A prioritizes supplying its spot goods, Supplier B replenishes its stock within 72 hours by adjusting its production line, and Supplier C serves as a secondary backup. Simultaneously, the system automatically generates priority confirmation requests to be sent to the collaborative platforms of these suppliers.

[0054] Application Results: This approach enabled rapid and optimal matching of supplier resources, enhancing the resilience and efficiency of the supply chain in responding to unforeseen events. This response model also strengthened the system's decision-making knowledge for similar emergency scenarios.

[0055] Example 4: Automated Optimization of Daily Warehouse Space Utilization Scenario description: During daily operation, the system continuously monitors the space utilization rate of each warehouse.

[0056] System application process: Edge computing processing: For high-frequency IoT sensor data in the warehouse (such as shelf weight and volume scanning), the edge computing capabilities of the heterogeneous data access gateway are aggregated locally, and only the daily warehouse utilization summary data is uploaded, reducing the pressure on the central system.

[0057] Automated decision-making and execution: When the system detects that the utilization rate of general hardware materials in a warehouse is consistently low, while the warehouse in another warehouse is short of the same type of materials, for such routine optimization with clear rules and low risk, the intelligent decision generation module automatically generates and executes a fine-tuning instruction to "balance the inventory of the two warehouses in the next routine transportation" under the supervision of a preset approval flow, and sends it directly to the warehouse management system.

[0058] Feedback optimization: The closed-loop feedback module tracks subsequent data to confirm that space utilization has improved, and fine-tunes the weight parameters of relevant indicators in the evaluation model, so that the system pays more attention to such efficiency points that can be automatically optimized.

[0059] Application effect: It realizes dynamic self-balancing of warehousing resources, freeing managers from tedious daily scheduling and allowing them to focus on more complex strategic decisions.

[0060] Example 5: Supplier Risk Precautionary Decision Making Based on Historical Knowledge Scenario Description: The system detected a slight fluctuation in the monthly delivery data of a key component supplier. The pattern was similar to the initial pattern recorded in the dynamic knowledge base a year ago when the supplier experienced delivery delays due to internal production scheduling issues.

[0061] System application process: Knowledge base matching: A dynamic knowledge base driven by closed-loop feedback and self-learning optimization modules automatically identifies current abnormal data patterns that have a high degree of similarity to historical cases.

[0062] Early Warning and Decision Recommendation: Before the evaluation model issues a formal deviation warning, the system alerts management to the risk and prioritizes recommending effective response strategies recorded in the knowledge base: "It is recommended to initiate an early communication mechanism with suppliers, verify their production schedules, and prepare to activate the pre-screening process for alternative suppliers." The recommended solutions are supported by historical case data.

[0063] Proactive intervention: Management adopted suggestions and conducted early communication, discovering that the supplier did indeed have a temporary capacity bottleneck, thus enabling early coordination and avoiding supply interruptions at the project site.

[0064] Application results: Transforming past individual experiences into reusable organizational wisdom has enabled a shift from passive response to proactive prevention, significantly enhancing the collective decision-making capabilities and resilience of the supply chain.

[0065] 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 power grid supply chain operation efficiency assessment and intelligent decision-making system integrating multi-source data, comprising a heterogeneous data access gateway, a data quality governance engine, and a dynamic semantic association module, characterized by: The heterogeneous data access gateway is configured to seamlessly access and collect in real time structured, semi-structured, and unstructured raw data streams from the power grid material master data management platform, supplier relationship management system, IoT device sensing network, logistics tracking system, warehouse environment monitoring system, and external macroeconomic and social data platform via a standardized adapter interface. The data quality governance engine has a built-in set of configurable integrity, consistency, and accuracy verification rules to automatically clean, repair, and label the incoming data streams and generate a data health metric report. The dynamic semantic association module is based on a preset power grid supply chain domain ontology model to intelligently identify, classify, and map relationships between the governed multi-source data entities, constructing a dynamic association data network with spatiotemporal labels, with material categories, suppliers, project requirements, warehousing nodes, and transportation units as core elements, providing a unified and high-quality data foundation for subsequent efficiency evaluation.

2. The power grid smart supply chain operation efficiency assessment and intelligent decision-making system integrating multi-source data as described in claim 1, characterized in that: Furthermore, a multi-dimensional evaluation model for supply chain operational efficiency is included. Based on the aforementioned dynamic relational data network, this model constructs an evaluation index system encompassing four dimensions: resource efficiency, process agility, collaborative response, and risk resilience. Resource efficiency is measured by calculating indicators such as inventory turnover rate, warehouse space utilization rate, and capital occupation cost. Process agility is evaluated by quantifying order fulfillment cycle, on-time delivery rate, and anomaly handling timeliness. Collaborative response is reflected by analyzing the accuracy of demand forecasting and the frequency and quality of information interaction on the supplier collaboration platform. Risk resilience is comprehensively determined by simulating the impact of external events on the critical path of the supply chain, assessing supplier concentration, and the preparedness of alternative solutions. The evaluation model employs an adaptive weight allocation algorithm based on an expert knowledge base to dynamically adjust the weights of each indicator and output a comprehensive efficiency index and a multi-dimensional radar chart, achieving a panoramic and quantifiable operational status diagnosis.

3. The power grid smart supply chain operation efficiency assessment and intelligent decision-making system integrating multi-source data as described in claim 2, characterized in that: The evaluation model incorporates a performance deviation calculation method based on a dynamic baseline. This method first establishes a dynamic and contextualized performance baseline for each evaluation indicator based on historical operating data, industry benchmark data, and real-time market demand fluctuations. The calculation module compares the current operating data with the dynamic baseline in real time, not only calculating the absolute deviation value, but also analyzing the trend, duration, and fluctuation amplitude of the deviation, using fuzzy logic to determine the severity level of the deviation, and locating the key data sources and business processes that cause the deviation. This elevates efficiency evaluation from a static numerical comparison to a dynamic trend diagnosis with early warning significance.

4. The power grid smart supply chain operation efficiency assessment and intelligent decision-making system integrating multi-source data as described in claim 3, characterized in that: Further, it includes a supply chain intelligent decision generation module. This module receives the efficiency deviation diagnosis results from the evaluation model and embeds a decision tree based on a rule engine and constraint satisfaction theory. This decision tree can automatically generate a set of priority optimization decision suggestions for different efficiency deviation scenarios. For example, when the inventory turnover rate of a specific material is diagnosed as significantly deviating from the benchmark, the decision module will integrate the current project requirements, in-transit material information, and supplier historical performance to generate multiple decision options, including "initiating supplier collaborative inventory replenishment", "adjusting safety stock parameters", and "recommending transfers from nearby warehouses", and estimate the potential impact of each option on the overall efficiency index to assist managers in making decisions.

5. The power grid smart supply chain operation efficiency assessment and intelligent decision-making system integrating multi-source data as described in claim 4, characterized in that: The intelligent decision generation module integrates a multi-objective optimization simulator. This simulator abstracts supply chain operations into an optimization problem encompassing multiple objectives, including cost, time, service level, and risk control. When making decision recommendations, the simulator allows users to set priority preferences for each objective. Subsequently, based on data within the system, it conducts large-scale, multi-round simulations of candidate decision schemes, quantifies and predicts the comprehensive impact of each scheme on the aforementioned multiple objectives under different market scenarios, and presents the simulation results in a visual manner. This achieves a leap from experience-based decision-making based on historical data to predictive decision-making based on future scenario simulations.

6. The power grid smart supply chain operation efficiency assessment and intelligent decision-making system integrating multi-source data as described in claim 5, characterized in that: Furthermore, it includes a closed-loop feedback and self-learning optimization module. This module continuously tracks the actual effect data of implemented decisions and compares it with the predicted effects of the decision generation module to form a feedback loop on the effectiveness of decisions. This module has a built-in causal analysis model to identify the strength of the causal relationship between decision actions and changes in efficiency indicators. Based on the analysis results, the module can automatically fine-tune the weight allocation algorithm in the evaluation model and the rule parameters in the decision module, so that the evaluation accuracy and the applicability of the decision recommendations of the entire system can continuously evolve and optimize themselves with the accumulation of operating time, without relying on external machine learning frameworks.

7. The power grid smart supply chain operation efficiency assessment and intelligent decision-making system integrating multi-source data as described in claim 1, characterized in that: The heterogeneous data access gateway has edge computing capabilities. It deploys lightweight data processing logic near the data source to perform preliminary filtering, aggregation, and feature extraction on high-frequency IoT sensor data. Only valuable event summaries or abnormal data are uploaded to the cloud center, which greatly reduces the data processing pressure on the central server, reduces network bandwidth requirements, and improves the response speed of the entire system for scenarios with extremely high real-time requirements.

8. The power grid smart supply chain operation efficiency assessment and intelligent decision-making system integrating multi-source data as described in claim 2, characterized in that: The visualization of the evaluation index system is achieved through an interactive digital twin cockpit. This cockpit constructs a virtual model that is fully mapped to the physical supply chain, and renders information such as resources, processes, and status in three dimensions. Users can not only intuitively view the efficiency index of each dimension, but also trace back from the macro comprehensive index to the micro raw data record through interactive operations such as drill-down, filtering, and clicking. This realizes the visibility, manageability, and traceability of the evaluation results, greatly improving the intuitiveness and penetrability of management decisions.

9. The power grid smart supply chain operation efficiency assessment and intelligent decision-making system integrating multi-source data as described in claim 4, characterized in that: The intelligent decision generation module is deeply integrated with the supply chain execution system and has hierarchical autonomous execution capabilities. For high-frequency routine decisions with clear rules and low risks, the system can directly convert decision suggestions into executable instructions under the supervision of a preset approval flow and send them to the procurement system or warehouse management system to achieve "unmanned" autonomous operation of specific processes. This frees managers from tedious tasks and allows them to focus on more complex strategic decisions and exception handling.

10. The smart power grid supply chain operation efficiency assessment and intelligent decision-making system integrating multi-source data as described in claim 6, characterized in that: The system also includes a dynamic knowledge base driven by the closed-loop feedback module. This knowledge base not only stores successful decision-making cases and their effect data, but also specifically includes abnormal patterns of performance deviation, fundamental analysis conclusions, and verified effective coping strategies. This knowledge is indexed and linked in a structured manner. When the system detects similar abnormal patterns again, it can prioritize recommending verified strategies from the knowledge base, along with historical cases as supporting evidence. This transforms individual experience into organizational wisdom, enhancing the collective decision-making ability and risk resistance resilience of the entire supply chain system.

Images