Power Assisted System Controller Design for Wheelchair

• Main Authors: Tsai, Yue-Ruen, 蔡岳潤
• Other Authors: Tseng, Chyuan-Yow
• Format: Others
• Language: zh-TW
• Published: 2018
• Online Access: http://ndltd.ncl.edu.tw/handle/5d39aq

碩士 === 國立屏東科技大學 === 車輛工程系所 === 106 === Power-assisted wheelchair (PAWC) is a power-assisted system installed on a manual wheelchair. It provides a calculated adequate auxiliary force to the wheelchair according to thrust value from the driver, so that users with poor strength can autonomously move to any road conditions such as slope and terrain. Since the user is usually an inconvenient or inferior elder, the performance indices of the system are stability, agility, safety and comfort to use, which must be achieved by properly integrating mechanical design and control techniques. The R&D team of Pingtung University of science and technology has completed a power-assisted mechanism, which has the advantages of quick installation, driving stability and agility. The mechanism is mounted on rear chassis of the wheelchair and is powered by a 24V 250W hub motor equipped with a 24V6.4Ah battery. This study is mainly focused on developing the control system and auxiliary force control technology for the power-assisted system. Issues include road slope estimate, driver’s thrust input estimation, and optimal auxiliary force design. The road slope estimation applies the method of Kalman filter combined with the complementary filter. With the real-time estimate of the road grade, an external force observer is designed to estimate the user's thrust input, then the load due to slope, the user's thrust input, the wheel speed, and the steering state of the wheelchair are taken into consideration to obtain a best auxiliary force, The power output of the auxiliary motor is controlled by closed current control loop to provide the user with an immediate, comfortable, and smooth operation. The work of this research can be divided into three parts: algorithm design, experimental system design and establishment, and finally performance verification and parameter calibration. Several tests have been conducted on various road conditions such as indoor narrow spaces drive, outdoor uphill, and running through bumps, etc. The test results showed the assistance force can quickly respond to user’s thrust input, and after the thrust is released, the assistance force can be gently lowered to move a short distance. Any uncomfortable force and misjudgment were not found to exist under the interference of the road irregular surfaces and obstacles. For the estimate of road slope, the test results showed that, on the road condition from level to 5 degrees uphill, the estimated slope smoothly converges to a steady state value within 3 seconds, with the steady-state error of less than 3.4% (the estimated average is 4.83 degrees). The steady-state error of the estimated thrust value was less than 11.25%, when on level road drive without the assistance power activated. For the case of power activated, the steady-state error of the estimated external force was less than 7.07%. In terms of overall performance, under the full load of 100kg and the use of 24V6.4Ah battery, the indoor driving range is averaged about 16km. For the effect of power assistance, the running distance per push action is about an average of 5 m under power assistance, which is 150% of the unassisted running distance of 2 m. When the auxiliary force is provided, all the values of acceleration are well controlled to less than 0.3m/s2, which conforms to the comfortable requirement of ISO2631-1 standard.