On December 20, 2018, the first tNetrol-NSI4/2-660-X image speed sensor developed by Ningbo Times was successfully tested on the magnetic levitation line of Tongji University.
In Ningbo Times, a non-contact image speed sensor based on spatial filtering principle was developed, which includes optical sensor and signal processing components. Through signal processing measures such as image acquisition and spatial filtering on rail surface, the information of train running speed and direction can be obtained, which has frequency resolution. It has the characteristics of high rate, high measurement accuracy and strong anti-interference ability. The research began in February 2018. After 10 months' efforts of the project team, technical problems such as insufficient light compensation performance, high fundamental frequency component and high frequency harmonic interference were solved. Besides the key technologies such as high-speed image acquisition, signal filtering and compensation, direction recognition were mastered. Finally, the product was developed and successfully passed the type test.
The sensor uses high-speed linear image device with a scanning frequency of 100 KHz and a measurement response speed better than 50 ms. Through high-speed signal spectrum analysis and feature extraction, it achieves a speedy resolution of 0.2 km/h within 100 km/h and 0.2% above 100 km/h. Based on the higher velocity resolution at low speed, the sensor can measure the displacement of the train in the process of entering and stopping. The resolution is better than 125 px. In addition, the sensor also has good electromagnetic protection design and output signal redundancy design, which can work stably and reliably in the harsh electromagnetic environment of maglev train.
With the development of high-speed track era, the requirement of real-time safety control for train speed sensor is becoming higher and higher. The development of image speed sensor fills the gap in the field of non-contact speed sensor of Ningbo times electric, and makes the sensor products take a solid step towards the direction of intelligence and platform-base.