Improved Positioning Method for Magnetic Encoder Type AGV using Extended Kalman Filter and Encoder Compensation Method

Improved Positioning Method for Magnetic Encoder Type AGV using Extended Kalman Filter and Encoder Compensation Method

This paper presents an improved positioning method for a Magnetic Encoder type Guided Vehicle(MEGV) using the Extended Kalman Filter and Encoder Compensation Method. The magnetic encoder system isone of several available guidance systems for autonomous guided vehicles using magnetic sticks that are buried atregular intervals (such as near landmarks, turning points, and work places) on designated paths. The system guidesMEGVs on a pre-defined path using either of two types of devices: encoders or magnetic positioning devices. Theencoder information is used in a range of positions between the magnetic sticks, and the magnetic positioning deviceis used to correct positioning of MEGV using global positioning of a magnetic stick. However, calculatingthe exact position of a MEGV is challenging because of errors (cumulative error of the encoder and disturbancesin the general magnetic field). Therefore, this study proposes a method, which is a combination of EKF and ECM,for positioning MEGVs. In the proposed method, EKF first estimates the position of the MEGV; then, ECM correctsthe error of the encoders. To analyze the performance of the proposed method, a MEGV was designed anddeveloped. The proposed method was compared with three other positioning methods (that use encoders, magneticencoders, or EKF), and experiments were performed under similar working conditions. The experimental resultsdemonstrated that the proposed method is superior to the other methods.

This paper presents an improved positioning method for a Magnetic Encoder type Guided Vehicle(MEGV) using the Extended Kalman Filter and Encoder Compensation Method. The magnetic encoder system isone of several available guidance systems for autonomous guided vehicles using magnetic sticks that are buried atregular intervals (such as near landmarks, turning points, and work places) on designated paths. The system guidesMEGVs on a pre-defined path using either of two types of devices: encoders or magnetic positioning devices. Theencoder information is used in a range of positions between the magnetic sticks, and the magnetic positioning deviceis used to correct positioning of MEGV using global positioning of a magnetic stick. However, calculatingthe exact position of a MEGV is challenging because of errors (cumulative error of the encoder and disturbancesin the general magnetic field). Therefore, this study proposes a method, which is a combination of EKF and ECM,for positioning MEGVs. In the proposed method, EKF first estimates the position of the MEGV; then, ECM correctsthe error of the encoders. To analyze the performance of the proposed method, a MEGV was designed anddeveloped. The proposed method was compared with three other positioning methods (that use encoders, magneticencoders, or EKF), and experiments were performed under similar working conditions. The experimental resultsdemonstrated that the proposed method is superior to the other methods.