SUMMARYIn this paper, we propose a biomechatronic design of an anthropomorphic artificial hand that is able to mimic the natural motion of human fingers. The prosthetic hand has 5 fingers and 15 joints, which are actuated by 5 embedded motors. Each finger has three phalanges that can fulfill flexion-extension movements independently. The thumb is specially designed to move along a cone surface when grasping, and the other four fingers are well developed based on the four-bar link mechanism to imitate the motion of the human finger. To accomplish the sophisticated control schemes, the fingers are equipped with numerous torque and position sensors. The mechanical parts, sensors, and motion control systems are integrated in the hand structure, and the motion of the hand can be controlled through electromyography (EMG) signals in real-time. A new concept for the sensory feedback system based on an electrical stimulator is also taken into account. The low-cost prosthetic hand is small in size (85% of the human hand), of low weight (420 g) and has a large grasp power (10 N on the fingertips), hence it has a dexterous and humanlike appearance. The performance of the prosthetic hand is validated in a clinical evaluation on transradial amputees.
In order to overcome the difficulty in selecting parameters of support vector machine (SVM) when modeling the PT fuel system fault diagnosis, SVM optimized by particle swarm optimization (PSO) algorithm was proposed. The PSO-SVM model was established and the fault multi-classifiers of the SVM were got. The pressure signal of the PT fuel inlet and outlet at different rotational speed and conditions was collected. The algorithm of PSO-SVM was used to train and recognize the pressure signal. The result of experiment confirms the validity of this method through comparison of the BP-NN, SVM and the PSO-SVM.
Due to the inaccuracy of reasoning conclusion because of the discrepancy among the cases ontology in the process of case reuse or revision, a new reasoning method for fault diagnosis based on ACO and CBR is proposed by this paper. This method uses CBR to reason new cases firstly, if it can match successfully brings to the corresponding results, else use ACO to reason the cases which have not matching in the case-library. As a result, the accuracy and efficiency of fault diagnosis are improved greatly and use the characteristic of the strong memory in CBR to repair the shortcoming of ACO reasoning method that can improve it capability. The method adopting fault cause-symptom matrix to describe the cases and case-library has many good characteristics of conciseness, convenience and extendibility.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.