Application of Multi-sensor Information Fusion Based on ARM in Industrial Control

0 Preface

Modern industrial production is characterized by comprehensive, complex, large-scale, continuous, and uses a large number of sensors to monitor and control the production process. The emergence of multi-sensor systems has led to a dramatic increase in the amount of information, and the use of information fusion technology can make more efficient use of information resources. In a complex industrial control system, the control process needs to involve multiple information at the same time, especially the relationship between the information, the information characteristics of the organic combination of information and the overall status of the information, and the process characteristics described according to the comprehensive situation. Take control.

1 Multi-sensor information fusion

1.1 The concept of multi-sensor information fusion

Multi-sensor information fusion refers to the coordinated use of multiple sensors, synthesizing the partial incomplete observations provided by multiple sensors distributed in different locations and the related information in the associated database, eliminating the redundancy that may exist between multi-sensor information. And contradictions, and complement each other, reduce its uncertainty, and obtain a process of describing the consistency of objects or the environment. That is, multi-level, multi-faceted, multi-level processing of data from multiple sensors, resulting in new and meaningful information that is not available with any single sensor.

1.2 Multi-sensor information fusion method

Commonly used information fusion methods are weighted average method, Kalman filter, classical reasoning method, Bayesian estimation, DS evidence decision theory, quality factor method, template method, entropy theory, cluster analysis, fuzzy reasoning, production rules, inheritance. Algorithm, neural network. Among them, the neural network method has strong information processing capability, and the neural network is a powerful tool for complex industrial intelligent monitoring and control systems and information processing uncertainty, thus providing a good information fusion. method.

The basic idea of ​​neural network for information fusion is to determine the classification criteria based on the sample similarity received by the current system. The determined method is mainly expressed in the distribution of network weights, and the neural network learning function can be used to acquire knowledge and obtain the uncertainty reasoning mechanism. A method of adaptive resonance theory ART (AdapTIve Resonance Theory) in a neural network is employed. Figure 1 is a network diagram of ART-2 that can process analog information.

The competitive learning mechanism and the self-stabilizing learning mechanism are the basis of the adaptive resonance theory (ART). The competitive learning mechanism only changes the weighting factors associated with the winners of the competition, while all other weighting factors remain unchanged. Through learning, the observation vector sets of different objects find their respective winning output components, so they can be naturally classified according to the winner's number.

The self-stable learning mechanism consists of information feedback channel, comparison mechanism and corresponding algorithm. The working process is described as follows: 1) competitive selection operation; 2) constitute feedback information channel; 3) compare similarity; 4) adjust weight coefficient.

2 Industrial control systems and information fusion

2.1 Information Fusion in Industrial Control

In industrial control, process operating conditions are referred to as operating conditions. In a simple system, the output of a sensor can generally reflect the operating conditions. In complex industrial production processes, the operating conditions cannot be directly represented by one or several process variables, and the output of a sensor is only a side of the description of the operating conditions. With appropriate information fusion methods, the fusion of multiple sensor information describing the operating conditions from different sides makes it possible to obtain a complete description of the operating conditions, which can be operated and real-time intervention, or the system can be automatically controlled according to working conditions. Figure 2 shows the principle of information fusion control, which can be broken down into the following parts:

1) Sensor information acquisition, including multi-sensor system, sensor information pre-processing and soft measurement, human-machine interface. This part includes hardware and related software. It should detect the information reflecting the running status and environment of the object as much as possible, including the state quantity, controlled quantity and environmental information of the object.

2) Clustering fusion control, composed of a series of software modules, is the core part of completing intelligent monitoring and control.

3) Interpretation agencies, including hardware and related software, such as pictures, text, sound, light, multimedia output devices. Interpret the current state of the system and the results of clustering fusion, and answer the questions raised by the user through the human-machine interface. In an intelligent monitoring system without automatic control function, such as fault diagnosis system, production operation guidance system, etc., the interpretation mechanism completes all output functions of the system: displaying the current working conditions of the production system, fault diagnosis results, hidden faults and dangers Suggestions and guidance for forecasting, production operations, etc.

4) The actuator, including the relevant power amplification and execution device hardware, implements automatic feedback control based on the results of the clustering fusion control operation.

2.2 system hardware design

The system is mainly composed of ARM920T core processor SamsungS3C2410, external RAM, Flash, D/A converter chip, LCD and RS232 interface. S3C2410 chip has an 8-channel 10-bit A/D converter inside, so there is no need to expand A/D. Conversion chip, the overall hardware block diagram of the system is shown in Figure 3.

2.3 System Software Design

1) Linux operating system porting

Linux has the characteristics of open source code. Although the Linux operating system is not a micro-kernel structure, it uses dynamic module loading, which makes Linux cut very convenient. The modular structure allows users to easily configure it and remove the user system. Modules that reduce system overhead. The Linux system includes a bootloader, a kernel, and a root file system. The system is developed under the Linux system. The steps include: (1) configuring the bootloader and the Linux kernel, and burning it to the target platform; (2) Burn the root file system; (3) establish a cross-compilation environment.

2) Clustering fusion subroutine algorithm

The clustering fusion subroutine is the core of the whole system, including the fusion database and expert knowledge base in information fusion, as well as the data and knowledge needed in the process of monitoring, control and operation, and transformation. 4 is a flow chart of a clustering fusion algorithm.

3) Interface design

Qt/Embedded is produced by Norwegian TrollTech and is a complete C++ application development framework, so Qt is used for design. Qt includes a class library and tools for cross-platform development. In addition to the C++ library and rich APIs, QT abstracts external input devices into input events, which can support a variety of hardware devices, enabling users to not only develop graphical interface programs, but also control the underlying hardware. Figure 5 is an information fusion control interface that simulates three sensor inputs for boiler water level, flow, and pressure.

3 Summary

Multi-sensor information can accurately, accurately and reliably reflect the characteristics of objects and environments, and the introduction of information fusion technology into industrial control can make more efficient use of information resources. As a 32-bit RISC microprocessor, ARM has the characteristics of high integration and high reliability. It can solve the problem of low reliability and poor anti-interference ability caused by the relative dispersion of equipment and devices in traditional industrial production processes.

Picture lamp is a kind of high-end flexible decorative lamp, characterized by low power consumption, long life, high brightness, easy to bend, maintenance free and so on. In addition, picture lights light up the whole work, and different lights can better show the emotions expressed in the work. Correct picture lamp is absolutely the pen that makes the finishing point, turn a piece of work from common to breathtaking, high quality picture lamp, as well as light ornament will bring different artistic atmosphere for you.

Picture Light

Photo Clip String Lights,Photo String Lights,Led Picture Light,Lighting Images

JIANGMEN LEDERLIGHT LIGHTING Co.,LTD , https://www.lederlightcn.com