SENSOR SYSTEM OPTIMIZATION OF A WALKING MACHINE WITH DYNAMIC STABILITY

Abstract

Walking machines can be more effective as compared with wheeled and tracked vehicles when they travel along a rough landscape as they come into contact with the ground at separate points and cause less damage to agricultural crops. Machines of this type are applicable in open ground and in protected greenhouse conditions for an accurate monitoring of the condition of cultivated plants and such negative growth factors as weeds, pests, and diseases. A walking machine has a lot of sensors, which may cause some connection difficulties due the limitations of microcontrollers or the method of data transfer. The paper discusses optimization methods for increasing the transfer data rate of two types of sensors with different data buses. The author presents connection diagrams of the Dallas Semiconductor DS18B20 temperature sensors and weight sensors based on the HX711 chip. The paper describes a developed bench for testing a walking machine and presents an experimental walking machine. It has been experimentally established that six digital lines with bit-by-bit conversion to a digital value are quite sufficient for reading logical values at 80 Hz speed with 25-bit data packet. As a result of optimizing data transfer of the on-board computer with a large number of sensors, the speed of polling sensors increased in 30 times (from 30 seconds to an interval of less than one second). The suggested parallel method of data reading for load sensors has provided for simultaneous scanning of six sensors at a speed of up to 80 Hz, and up to 12 Hz for sequential scanning.