A Low-Cost Intelligent Gas Sensing Device for Military Applications

Abstract

The field of electronic noses and gas sensing has been developing rapidly since the introduction of the silicon based sensors. There are numerous systems that can detect and indicate the level of a specific gas. We introduce here a system that is low power, small and cheap enough to be used in mobile robotic platforms while still being accurate and reliable enough for confident use. The design is based around a small circuit board mounted in a plastic case with holes to allow the sensors to protrude through the top and allow the natural flow of gas evenly across them. The main control board consists of a microcontroller PCB with surface mount components for low cost and power consumption. The firmware of the device is based on an algorithm that uses an Artificial Neural Network (ANN) which receives input from an array of gas sensors. The various sensors feeding the ANN allow the microcontroller to determine the gas type and quantity. The Testing of the device involves the training of the ANN with a number of different target gases to determine the weightings for the ANN. Accuracy and reliability of the ANN is validated through testing in a specific gas filled environment.

Review on Smart Gas Sensing Technology

“…The error backpropagation (BP) algorithm has been proposed to solve this problem [118]. The main principle of BP is to adjust network weights and thresholds through error feedback [119,120]. The BP algorithm requirs a long training time.…”

“…ANN [91,119,120,121,122,123,124,125,126] is a universal method for prediction and analysis of complex gas which sufficient approximates the nonlinear relation and has the ability of associative learning. Casey et al [91] use low-cost sensors to quantify 4 trace gas species in an oil and gas production region, and confirm ANN can represent the complex nonlinear behavior in response to the ambient environment among multiple gas sensors, including temperature, humidity, and atmospheric chemistry.…”

A robust spike-based gas identification technique for SnO<inf>2</inf> gas sensors

“…When the sensor is exposed to gas, the sensor resistance drops according to (1). Readout is performed when the resistance becomes steady.…”

“…Gas identification systems are widely used in military, industry and medical applications [1]- [3]. Examples include combustible gas detection in mine sites and toxic gas leakage in laboratory.…”