Deep learning-based energy-response production analysis system and method for automotive-grade electrical components

Abstract

The invention provides an automobile-grade electrical element energy response type production and analysis system based on deep learning and a method thereof, the system comprises a manually-driven intelligent manufacturing and execution module, a deep search module and an intelligent photovoltaic and clean room control module, and through MES and EMS integration standard establishment and improvement and deep combination of an industrial LLM and a knowledge graph, the energy response type production and analysis of the automobile-grade electrical element is realized. A highly adaptive, intelligent and sustainable manufacturing environment is created, optimized production and management of a man-machine cooperation mode are realized, proposed energy response manufacturing execution and analysis are realized, a production process is monitored and managed in real time, and energy availability and operation requirements are dynamically responded, so that the production quality and the production efficiency of automobile-grade electrical elements are improved.

Description

Abstract The present invention discloses a deep learning-based energy-response production analysis system and method for automotive-grade electrical components. The system includes a manually driven intelligent manufacturing and execution module, a deep search module, and an intelligent photovoltaic and clean room control module. By establishing and perfecting the integration standards of MES and EMS, and deeply integrating industrial large language models with knowledge graphs, a highly adaptive, intelligent and sustainable manufacturing environment is constructed. The invention realizes optimized production and management under the human-machine collaboration mode, achieves the proposed energy-responsive manufacturing execution and analysis, monitors and manages the production process in real time, and dynamically responds to energy availability and operational requirements, so as to improve the production quality and production efficiency of automotive-grade electrical components.

Claims

1. A deep learning based automotive grade electrical component energy response based production and analysis system for creating a highly adaptive, intelligent and sustainable manufacturing environment, characterized in that, The system comprises: An artificially driven intelligent manufacturing and execution module for generating fault or abnormality alarms, a dashboard displaying real-time production activities of all production lines, raw material inventory and finished or in-process product status, just-in-time production planning, pre-alarm of all materials about to expire, retrieval of appropriate production files, real-time input data for just-in-time production, document creation and usage of production machines or equipment; A deep search module for internally generating fault or abnormality alarms sending alarm messages to the management group, generating manufacturing reports, summarizing operational data, providing actionable recommendations and cross-department inquiries; An intelligent photovoltaic and clean room control module for generating fault or abnormality alarms sending alarm messages to the management group when detecting the presence of smoke and high temperature, or when the intelligent photovoltaic system has high / low voltage, high / low current or high temperature, and when the room temperature and humidity in the clean room are high, or the PM / TVOC value is high; and controlling the intelligent photovoltaic system to output at rated power and generate carbon emission reduction data; while controlling the temperature range in the corresponding production environment of the clean room to be between 18-28℃, the humidity range to be between 45-65%, and the detection of air quality index to cover PM1.0, PM2.5 and TVOC.

2. The deep learning-based automotive-grade electrical component energy response production and analysis system according to claim 1, characterized by, The artificially driven intelligent manufacturing and execution module comprises an artificial intelligence driven manufacturing execution system, a warehouse management system software platform, an Internet of Things system software platform, an Internet of Things sensor network, an electronic work instruction book hardware system and a server, an intelligent wired network terminal and Wi-Fi network; The artificial intelligence driven manufacturing execution system and the warehouse management system software platform are used to realize the processes of plan management, production execution, quality inspection, raw material warehouse, finished product warehouse and equipment management; The Internet of Things system software platform is used to support the connection of at least 150 points; The Internet of Things sensor network predicts quality control in the production process according to temperature, smoke or particle monitoring; The electronic work instruction book hardware system and the server support the connection communication of at least 200 contacts through the intelligent wired network terminal and Wi-Fi network. 3.The deep learning-based automotive-grade electrical component energy response production and analysis system of claim 1, wherein The deep search module comprises an internal server, at least 4TB of hard disk storage, a network switch and a converter. The internal server is used for deep search of R1 inference model and distributed computing through the network switch and the converter, and generates fault or abnormality alarms sending alarm messages to the management group, generates manufacturing reports, summarizes operational data, provides actionable recommendations and cross-department inquiries.

4. The deep learning-based automotive-grade electrical component energy response production and analysis system according to claim 1, characterized by, The intelligent photovoltaic and clean room control module comprises an intelligent photovoltaic control system, a photovoltaic mechanical frame, an artificial intelligence optimized clean room and a smart transformer / grid; The intelligent photovoltaic control system coordinates photovoltaic output and manufacturing execution system production planning through a controller, and supports artificial intelligence grid management, fault / temperature diagnosis and remote thermal imaging; The photovoltaic mechanical frame is used to integrate multiple sensors; The clean room is linked with the signals of real-time energy availability / virtual power plant for control; The smart transformer / grid is used for testing of bidirectional energy flow between the smart photovoltaic control system, clean room, and production load.

5. The production management method of the automotive-grade electrical component energy response type production and analysis system based on deep learning according to any one of claims 1 to 4, characterized by, The method comprises the following steps: S1, according to the MES / document control system, the latest production drawings are obtained for production post personnel to use; S2, production post personnel produce automobile electrical components according to material composition for production LOT batch number and sample plate control; S3, whether the product is qualified is checked, and after the unqualified product is checked, the number of adjustment or material replacement needs to be rectified; S4, after the qualified product is checked, the employee / machine reports on duty, and first, middle, and end piece inspection is performed, and production stop alarm is produced; S5, when the checked product is unqualified, production is stopped, and the product / material is isolated for rectification; S6, then, an abnormal report C.A.R is automatically generated, and is pushed to the supervisor for rectification and proper handling, after rectification, the supervisor must fill in the problem source and corrective measures, and after confirmation by a third party, the abnormal report is closed; S7, then, information is pushed to the C.E.O and senior management; S8, when the first, middle, and end piece inspection in step S4 is qualified, the employee / machine reports production number efficiency monitoring at regular intervals, through AI learning, target is formulated, production quantity is completed to each employee / machine, and employee / machine performance score is evaluated, the MES system sets the upper limit value of the scrap rate, and if the scrap rate exceeds the upper limit, an abnormal report is immediately issued to stop production and isolate the product; S9, when the product inspection in step S8 is unqualified, production is stopped, and the product / material is isolated for rectification; return to step S3 to rectify the number of adjustment or replace the material, and at the same time, the inspection situation is fed back to step S6 and step S7 in turn; S10, when the product inspection in step S8 is qualified, production is completed; the finished product is sent to OQC detection; S11, unqualified products are checked, and returned to step S8 for return for re-detection, or returned to step S2 for return for rework; S12, after the qualified product is checked, it is transferred to the finished product warehouse for packaging and delivery.

Images