该系统采用三块TCS230颜色传感器，监测轮船经过后海水的颜色变化的情况，数据采用就地处理，并运用改进型BP神经网络算法进行污染模式训练和模式识别。通过nRF401无线数字收发器，把污染识别的结果发送回监

视器，再通过字符和语音对识别结果进行报告。把轮船污染数据采集器安装在轮船尾部两侧进行实时监测，结果表明，监测速快，精确度高。-The system uses three TCS230

color sensor to monitor the ship through the water after the color changes， the data used in situ processing， and using improved BP neural network algorithm training and pattem recognition of pollution pattems． By nRF401 wireless digital

transceiver， the pollution， the results of recognition sent back to the monitor， and through the character and voice of the recognition results to report． The ship pollution data collector installed on ships on both sides of the tail of real－time monitoring results show that the monitoring of speed and high accuracy．