The Design And Implementation Of3D Electronic Compass Of Photogrammetric System

According to the geomagnetic field measured value and the measured value of the acceleration, three-dimensional electronic compass is a device to calculate the three-dimensional attitude, it is available for real-time measurement of the heading angle, pitch angle and roll angle of carrier platform. Three-dimensional electronic compass has a simple structure, low cost, ease of integration, high precision, etc. Because of its cost-effective, electronic compass is widely used in attitude measurement of the carrier platform and integrated navigation system. However, the geomagnetic field is very weak measured by three-dimensional electronic compass, the compass is very susceptible to various types of electromagnetic interference that directly affects the measurement accuracy of the compass. Therefore, error factors caused by the three-dimensional electronic compass are analyzed and processed, and reasonable electronic devices and filtering algorithms are selected to design the compass, improving the accuracy of three-dimensional electronic compass is the purpose of this paper.The measurement principle and basic structure of three-dimensional electronic compass are studied in this paper. Besides, the factors that affect the measurement accuracy also be analyzed, then a triaxial accelerometer, three-axis magnetometer sensor and microprocessor are used to design compass. According to the electromagnetic sensitive characteristics of three-dimensional electronic compass, the theories of signal integrity, rational distribution, wiring, setting filtering measures, making the necessary electromagnetic shielding design, its anti-jamming capability is enhanced. Hardware can effectively inhibit the partial interference, but it can not completely eliminate the noise in the complex environment of the three-dimensional electronic compass. In order to solve the interference, process is done in software level. In this paper use the measured values filtered by the median value of the average filter the random interference to reduce random signal interference. Using the least squares method to correct the acceleration sensor error of the measurement results, then makes it more nearer to the real value. At the same time, neural network compensation Kalman filter algorithm is introduced, excellent modeling capabilities of the neural network and the excellent ability to estimate of Kalman filtering are combined, the results of the Kalman filter are compensated by using neural networks. Training the neural network by ’Standard’ data, sending filtered data by Kalman filter to the neural network which had learn, to optimize the measurement data. Finally, in a real environment combined with electronic theodolite verified the experiments, the analysed data confirmed the accuracy of the neural network compensation Kalman filter algorithm, and shows that the design of the three-dimensional electronic compass accuracy is slightly higher than the domestic product.