Factory Energy Management System Integrating Artificial Intelligence

Abstract

An AI-integrated factory energy management system deeply integrates the Product Lifecycle Management System (PLM), Manufacturing Execution System (MES), and Energy Management System (EMS) through a standardized application programming interface (RESTful API). It unifies data through a cloud data platform and deploys an AI engine module to achieve intelligent functions such as predicting factory energy consumption, optimizing scheduling, and detecting anomalies through energy prediction models, scheduling optimization models, and anomaly detection models.

Claims

1. A factory energy management system integrating artificial intelligence, comprising: a product lifecycle management module for managing product bill of materials, process paths, and design change events; a production execution management module for managing work order scheduling, batch production, and equipment operation; an energy management module for monitoring real-time consumption of electricity, water, and gas and calculating carbon emissions; and an application interface module connecting the product lifecycle management module, the production execution management module, and the energy management module, wherein the application interface module communicates via expressive state transmission (RE). The STful protocol enables bidirectional real-time data exchange between the aforementioned three modules; a cloud data platform, which is connected to the application interface module, is used to unify the data of the product lifecycle management module, the production execution management module, and the energy management module. The cloud data platform includes a time-series database and a distributed database; and an artificial intelligence engine module, which is connected to the cloud data platform to read the unified data from the cloud data platform and perform machine learning based on the data from the time-series database and the distributed database.

2. The factory energy management system integrating artificial intelligence as described in claim 1, wherein, It also includes an application output module, which is connected to the artificial intelligence engine module. The application output module displays real-time energy consumption trends, optimal scheduling suggestions, electricity cost simulation calculations, and outputs equipment anomaly notifications and root cause analysis reports. It also supports product energy cost tracking and design and manufacturing energy consumption gap analysis.

3. The factory energy management system integrating artificial intelligence as described in claim 1, wherein, The artificial intelligence engine module includes an energy prediction model, which adopts a combination architecture of long short-term memory neural network (LSTM) and gradient boosting algorithm (XGBoost). It takes process type, equipment status, historical energy consumption data and ambient temperature as input features and outputs the predicted kWh consumption and confidence interval for each batch.

4. The factory energy management system integrating artificial intelligence as described in claim 1, wherein, The artificial intelligence engine module includes a scheduling optimization model, which adopts a hybrid optimization architecture combining reinforcement learning and genetic algorithm. It takes the electricity price time schedule, order urgency and equipment availability as inputs, minimizes electricity costs and peak load as the objective function, and outputs the best production scheduling solution.

5. The factory energy management system integrating artificial intelligence as described in claim 1, wherein, The artificial intelligence engine module includes an anomaly detection model that uses an autoencoder and isolated forest algorithm to detect anomalies in real-time current data, power factor and equipment temperature, and automatically trigger cross-system alarms and root cause analysis reports.

6. The factory energy management system integrating artificial intelligence as described in claim 1, wherein, The time series database stores complex time series electricity data, while the distributed database stores complex historical energy consumption analysis data.

Images