Design and implementation of vehicle-mounted positioning and orientation navigation system detector

1. Introduction

The vehicle-mounted positioning and orientation navigation system refers to a device that is loaded on wheeled and tracked vehicles and can autonomously provide the vehicle with azimuth reference and position information. The vehicle-mounted positioning and orientation navigation system described in this paper consists of a north finder, a bearing holding instrument, an odometer, an altimeter, etc. Among them, the north finder and the bearing holding instrument are extremely important parts. Their errors or failures directly affect the orientation and positioning accuracy of the system, and may even cause serious consequences. Therefore, it is particularly important to accurately detect and diagnose their faults. At present, the system adopts the traditional fault detection mode, which compares the difference between the actual value and the reference value of the relevant signal of each module through hardware and software. If it exceeds the allowable range, it is considered to be abnormal or faulty, and its accuracy and controllability are poor. In addition, the system has the following two shortcomings in performance detection and fault diagnosis: First, due to the constraints of volume and weight, the system design does not reserve interfaces with external detection instruments (such as oscilloscopes, multimeters, etc.), so it is impossible to perform maintenance in the field environment; second, the system cannot save and display the data of the working status of the internal inertial device at each period, and these data are extremely important for the use, management and maintenance of the system.

2. Basic idea

Obtaining the status data of the inertial device in the positioning and orientation navigation system during operation is a prerequisite for its performance analysis and fault diagnosis. Since the navigation system itself does not have a reserved detection interface, conventional detection instruments are obviously powerless. However, we noticed that the navigation system not only has a dedicated A/D module, V/F (voltage/frequency) conversion module and computer module responsible for collecting inertial device signals, but also has a reserved RS-232 serial communication port. Inspired by this, we designed a detection device - a detector, which can achieve the performance test and fault diagnosis of the navigation system while keeping the hardware structure of the navigation system unchanged and the performance indicators unaffected. To this end, we only need to make some simple additions to the navigation system software to add a working mode of the state to be detected on the basis of the original single positioning and orientation navigation state working mode. The detector communicates with the navigation system in a serial communication mode through the RS-232 serial port to obtain the data collected by the A/D module and V/F module in the system. The data processing and analysis method based on the BP neural network is adopted to make the performance detection and fault diagnosis of the navigation system accurate and convenient.

The working process of the detector is shown in Figure 1. When the navigation system is in the state of waiting for detection, the detector sends various detection commands to the navigation system through the serial port. The navigation system then enters the corresponding detection state and transmits the detection data to the detector. The detector completes the data processing and system performance analysis. After the detection is completed, the system automatically returns to the navigation state. During the detection process, the navigation computer receives commands, collects data, and sends data through interrupt control. After the operator issues various detection commands through the detector, the detection process is automatically completed by the computer without human intervention.

3. Hardware Configuration

As a special testing equipment, the detector is required to be able to easily connect with the navigation system and perform performance testing and fault diagnosis at any time and place where the vehicle-mounted positioning and navigation system needs to be tested and repaired. Therefore, the detector must have the characteristics of small size, light weight, low power consumption, high reliability, easy to carry, etc., and can be used in various harsh environments, especially in the field. According to the above requirements, the detector uses an embedded computer PC/104 module. The hardware configuration of the entire detector is shown in Figure 2.

The PC/104 computer module is the core part of the detector. During the detection process, the detector not only needs to complete the task of sending and receiving a large amount of data in real time, but also needs to perform complex processing, analysis and storage of various acquired data in real time, so it is required to have the highest possible performance. The PC/104 computer module we selected has a high-speed embedded Pentium CPU with a built-in floating-point arithmetic coprocessor and two 16C550 equivalent high-speed serial ports, which can quickly and reliably perform RS-232 serial communication and quickly and accurately complete the assigned functions.

The power module provides +5V and +12V power to the PC/104 computer module and the LCD display. To facilitate data exchange with other computers, the tester also has an Ethernet port.

4. Software Design

The software of the detector is designed based on Visual Basic 6.0, Visual C++ 6.0, Matlab6.5 and Access 2000 on the Windows platform. Visual Basic provides a powerful and easy-to-use serial communication control "MSComm", which has unique advantages in the programming language for developing Windows applications, but its functions in data processing are relatively weak, and Visual C++ 6.0 and Matlab6.5 can make up for its shortcomings. In terms of database operation, the detection software uses the data interface ActiveX Data Access Object (DAO) provided by Visual Basic to manage the database. DAO supports simple and convenient access to the Access database and is an easy-to-use application interface.

4.1 Communication Protocol

The transmission baud rate between the detector and the navigation system is 9600 bits/second, and the transmission frame format is: 1 start bit, 8 data bits, 1 stop bit, 1 even parity bit, a total of 11 bits of data. When the two parties communicate, they only need to add a start discrimination byte to the beginning of the pre-transmitted data. The communication process between the detector and the navigation system is: the detector sends a detection command - the navigation computer sends back relevant data.

4.2 Design of functional modules

In order to facilitate development, debugging, upgrading and maintenance, the software adopts a modular design concept. The entire software is mainly composed of four windows and six standard modules. The windows include display windows, function selection windows, navigation windows and detection windows; standard modules are functional modules for the detector to complete various tasks, including CRC check code calculation module, communication module, coordinate transformation module, data splitting module, database management module and detection data processing module. The CRC check code calculation module is used to generate serial communication check codes; the communication module uses the "input" and "output" properties and "OnComm" events of the serial communication control "MSComm" in Visual Basic to complete the data sending and receiving tasks; the coordinate transformation module is responsible for converting 54 coordinates into geocentric longitude and latitude coordinates; the data splitting module can split integer data into byte data for easy transmission through the serial port; the database management module saves the detected data in the database as needed; the detection data processing module is the core of the entire software, and gives corresponding processing results according to the type of detection data.

4.3 Processing of test data

In order to accurately judge the working status of the navigation system, we designed a fault diagnosis method based on BP neural network, and the action function selected S-type function. Since the signals of the gyroscope and accelerometer in the north finder are periodic signals when they are working normally, these signals can be used for performance detection and fault diagnosis using the same network. During the detection, a signal of one period is obtained, which is first passed through a low-pass digital filter to reduce and eliminate the influence of interference and noise in the detection data, and then normalized. After multiple simulation tests on the test data, we extracted the three quantities of peak value (MAX), kurtosis (KUR), and standard deviation (STD) as the input of the network. The network adopts the method of offline training and online use. The structure of the neural network is shown in Figure 3, where the number of neurons in the hidden layer, the weights and thresholds on each neuron are all obtained by training a large number of test sample data. Considering that the action function of the neuron is selected as the S-type function, the output training sample of the network is selected as 0.9 (faulty) or 0.1 (no fault). The neural network learning algorithm adopts an adaptive learning rate gradient descent back propagation algorithm, and each step of parameter update not only considers the current gradient direction, but also the gradient direction of the previous moment, thereby reducing the sensitivity of network performance to parameter adjustment and effectively suppressing the local minimum problem of training results. For a trained BP network, the output (out) of the network can be used to determine whether the corresponding inertial component is working properly. The test results show that the fault discrimination method based on the BP neural network is suitable for the fault diagnosis of inertial components of the vehicle-mounted positioning and orientation navigation system with high accuracy.

5. Conclusion

The successful development of the detector has solved the problem that the vehicle-mounted positioning and orientation navigation system cannot perform performance testing and fault diagnosis in the field environment. It can not only quickly and accurately obtain the status information data of the inertial components of the navigation system and complete the performance analysis and fault diagnosis of the system, but also ensure the normal operation of the navigation system and the performance indicators are not affected. Its working principle provides a new idea for the performance testing and fault diagnosis of existing equipment.