Customer Background

A power-generation equipment manufacturer is one of China’s earliest bases for developing generating equipment, and one of the important state-owned backbone enterprises under central management that relate to national security and the lifelines of the national economy. It has actively driven a new leap in China’s generating-equipment manufacturing level and independent innovation capability, forming leading products in nuclear, hydro, coal, gas, marine power plants, electric drives, and turnkey power-plant projects, with core technical capability at an advanced world level. To date it has produced 400 million kW of generating equipment, equipped more than 500 power plants at home and abroad, and exported to more than 40 countries and regions in Asia, Africa, Europe, and South America.

Customer Requirements

As traditional manufacturing transforms toward service-oriented manufacturing, demand for equipment monitoring and predictive maintenance has followed. Online monitoring and preventive maintenance of equipment itself can significantly raise inherent equipment performance, reduce unexpected downtime, keep equipment at optimal operation, and provide better service to customers. Beyond O&M optimization, collecting operating data can help or guide design departments to improve product design, fundamentally closing the loop from product design to full-lifecycle management. As production equipment is exported at scale, intelligent monitoring and predictive maintenance for operating equipment will also strongly support O&M capability on overseas projects, raising assurance while improving the performance of the whole equipment or solution.

Main technical requirements of the project are:

1. Closely combine industrial equipment, IoT, big data, and artificial intelligence to provide industrial enterprises with a massive-data storage and intelligent-analysis application platform that meets sensor data collection and real-time data storage and conversion, with rich interfaces and an extensible development framework, building business applications flexibly with data at the core.

2. Integrate and encapsulate industrial application components to build an enterprise/industry ecosystem: the platform includes rich mechanism-analysis and machine-learning modeling components, fully covering data preprocessing, feature engineering, model training, model evaluation, and other modeling stages, converting enterprise and industry modeling experience into platform algorithm components, lowering the modeling threshold and raising modeling results.

3. Build a full-lifecycle development framework for industrial intelligent apps: provide a model-based full-lifecycle framework for service development, deployment, and application monitoring, forming a platform innovation ecosystem centered on industrial app development, and providing customized, highly reliable, scalable industrial apps or solutions based on industrial microservices, forming a platform application ecosystem centered on value mining.

4. Establish enterprise and industry standards for next-generation generating-equipment intelligent application development: form end-to-end standards for generating-equipment intelligent application model development, data-interface specifications, model training, and model deployment, promoting standardized implementation of intelligent generating equipment and smart-power-plant application services.

5. Build a basic runtime and business-application environment for digital-twin applications: based on the industrial intelligent application development framework, build digital-twin models and application runtime environments for intelligent generating equipment and smart power plants. Using digital-twin models that fuse equipment mechanisms and multidisciplinary simulation, support digital-twin solutions for key scenarios such as product O&M and design-simulation iteration, power-plant O&M process modeling and optimization, and equipment-fault reproduction and maintenance-effect optimization.

Solution

The industrial intelligent application development framework (Realpower platform) provides a full-lifecycle application development framework for industrial intelligence. With models and components as the core design idea, the platform deeply fuses industrial mechanisms, industrial data, industrial algorithms, and enterprise knowledge, encapsulates them as models, and reuses them strategically, building an independent, open industrial-app development framework. It supports intelligent fault diagnosis and prediction, equipment health management, and system O&M optimization, while precipitating and encapsulating professional knowledge as the enterprise’s core assets.

Solution Advantages

1. Unified application-modeling technology based on industrial application scenarios

Unlike other analysis platforms, the wind-turbine fault-prediction and health-management platform strongly guides users in analyzing industrial problems, providing expert-guided analysis templates and a unified application-development environment for general-purpose components, equipment, and systems. The platform solidifies analysis methods from many industries into industry templates and analysis components, minimizing learning cost so users can focus on the business layer, quickly become familiar with the modeling environment, and complete modeling with maximum automation.

2. A SaaS application platform based on a microservice architecture

Independently develop and manage business application services around business-domain applications. By decomposing applications and services into smaller, loosely coupled components, deployment, management, and service-function delivery become simpler, building a business-enablement platform for the enterprise.

3. Building an equipment-fault model library based on equipment ontologies and AI

Use ontological knowledge to build an equipment-fault model library. AI reasoning helps maintenance personnel analyze fault phenomena based on existing fault-domain knowledge, find repair methods, accurately discover fault causes, and find solutions.

4. An overall industrial-internet solution for IoT + big-data applications

Provide an equipment-data collection and big-data storage and analysis platform, completing efficient storage and computational analysis of massive industrial data such as equipment operating data, environmental data, and maintenance data, realizing accumulation and use of enterprise knowledge assets.

Customer Benefits

By deploying the model-based industrial intelligent application development framework, the customer gained standardized, componentized support for smart-power-plant intelligent application development and O&M, helping the enterprise quickly build a unified business-development and service-runtime platform for smart power plants and intelligent generating equipment. It deepens new service-oriented manufacturing models such as big-data collection and analysis and remote O&M, and raises intelligent service capabilities in monitoring and tracing, predictive maintenance, quality control, supply-chain forecasting, target-customer credit assessment, and risk control.