FPGA-based Wireless Monitoring System Design Fall

With the extensive application of integrated circuits and sensor technology, we can monitor our own and living environment more handily. Due to the tendency of aging population, the health of the elderly have been paid much attention. Because their physical function decline and their adaptability and resilience decrease, the elderly are more prone to fall, and the injuries caused by falling are hard to recover, which will lead to permanent damage and even to death, if there is no immediate treatment. so the elderly fall monitoring very practical value. Therefore, there is a enormous significance and very high practical value to design a wireless falling monitoring system with low cost and high performance.In order to solve the problem, we combine the research of smart home system, and present a wireless falling monitoring system based on FPGA with the consideration of system’s functionality, practicability and other factors. The system consists of four parts: data acquisition section, wireless transmission section, initiative help-seeking section and falling identification section. The data acquisition section is implemented by triaxial accelerometer MMA7455gathering the acceleration data, in which the acceleration sensors are driven by the Zigbee CC2530chip, and the C language is adopted to write the driving program in the IAR software. The wireless transmission section adopts Zigbee CC2530chip’s protocol stack to send and receive data, and the application programs of the protocol stack are written in the IAR software. In the initiative help-seeking section, we design the keypads on the data transmitter to send the signals for help, whose function is verified by monitoring the keypad signals in different locations. The verified experiments are conducted30times. In the falling identification section, SVM classification algorithm is used to identify falling events. A NiosII soft core built by Quartus Ⅱ’s SOPC Builder on the FPGA chip, is designed to run the algorithm in the hardware unit of the falling identification section. And the software unit falling identification section is realized by using C language to compiler SVM classification algorithm on NiosII IDE. In the part of experimental verification, we adopt45sets of data to train the SVM classifier and find out the optimal interface. And then we conduct40experiments of falling verification, in which there are24cases falling events in6different ways. The experimental results are shown by the digital tubes and the buzzer.The Verifying of initiative help-seeking section is to monitor keypad signals in different locations. The experimental results show that the digital tube can display correctly the participants’number, and the buzzer sounds a reminder. After using multiple sets of data for SVM classifier is trained to identify10times experiments to verify the effectiveness of the fall detection section, which contains six cases fall events. Identify accuracy was90%, which meets the design requirements, the application can offer a new guarantee for the elderly health.After Experimental Verification, the detection rate and the whole identified accuracy are83.3%and92.2%respectively. The experimental results show that the wireless falling monitoring system designed in this paper can achieve the detection of falling movements and have good application values, which can be employed in falling susceptive field, such as gerocomium and hospital. Meanwhile, the system has good expansibility and can be used to construct the smart home system.