Multi-input Convolutional Neural Network Fault Diagnosis Algorithm Based on the Hydraulic Pump

Convolutional neural network used in fault diagnosis can effectively extract fault features in vibration signals. However, in the feature extraction of mechanical fault diagnosis, usually more than two feature signals including at least axial and radial vibration signals can be extracted. This paper proposes two multi-input convolutional neural network models based on the fault data of the aircraft hydraulic pump including axial and radial vibration. The first is the Independent Input Multi-input Convolutional Neural Network model. The two inputs are respectively used for convolution pooling operation with CNN, and are combined through the concatenate function before the fully connected layer, and then all frames are integrated and flattened by the flatten function. A one-dimensional array, finally enters the fully connected layer and outputs the result through the softmax function. The second is the Combined Input Multiinput Convolutional Neural Network, that is, combine two one-dimensional signals into a twodimensional signal in the input layer of the convolutional neural network and then perform convolution pooling, and finally output the result through the softmax function. The results show that the two models have good accuracy and stability, and the second one has a higher convergence and fitting efficiency than the first one.

Development of a Novel 2-DOF Rotary–Linear Piezoelectric Actuator Operating under Hybrid Bending–Radial Vibration Mode

The paper presents a numerical and experimental investigation of a novel two degrees of freedom (2-DOF) piezoelectric actuator that can generate rotary motion of the sphere-shaped rotor as well as induce planar motion of the flat stage. The actuator has a small size and simple design and can be integrated into a printed circuit board (PCB). The application field of the actuator is small-dimensional and high-precision positioning systems. The piezoelectric actuator comprises three rectangular bimorph plates joined with arcs and arranged by an angle of 120 degrees. A high-stiffness rod is glued on the top surface of each bimorph plate and is used to rotate the rotor or move flat stage employing contact friction force. Three U-shaped structures are used for the actuator clamping. 2-DOF rotational or planar movement is obtained by applying a harmonic or asymmetric electrical signal. The operation principle of the actuator is based on the superposition of the B20 out-of-plane bending mode of the bimorph plates and the B03 radial vibration mode of the ring. Design optimization has been performed to maximize amplitudes of contact point vibration. A prototype of the actuator was made, and a maximum rotation speed of 795.15 RPM was achieved while preload of 546.03 mN was applied. The linear velocity of 36.45 mm/s was obtained at the same preload force. Resolution measurement showed that the actuator can achieve an angular resolution of 17.48 µrad and a linear resolution of 2.75 µm.

An estimation algorithm for tire wear using intelligent tire concept

Real-time monitoring of tire wear is a hot spot in the research of automobile tires, and it has a great significance to ensure the safety of automobile driving. A tire wear estimation algorithm was proposed based on the relevant knowledge of finite element modal analysis theory and the concept of intelligent tires in this paper. First, the finite element model of the 205/55/R16 radial tire was established through the ABAQUS software, then the finite element method was used to simulate and analyze the influence of tire inflation pressure, load, tire wear, and speed on the tire radial vibration frequency. The simulation results show that inflation pressure and tire wear shows an upward trend with the increase of the vibration frequency of each order in the tire radial direction, and load and speed increase with what increases of tire radial increase frequency. Based on simulation analysis data, combined with the relationship between tire inflation pressure, load, tire wear, speed, and radial vibration frequency, a neural network-based tire wear estimation algorithm is proposed. The estimate results show that the predicted wear curve and the actual wear curve have a higher degree of overlap, the average error is 0.0874 mm, and the average error percentage is 2.78%, Thus, a feasible tire wear estimation algorithm is proposed.