Search results

Rotating machinery is a crucial component of large equipment, and detecting faults in it accurately is critical for reliable operation. Although fault diagnosis methods based on deep learning have been significantly developed, the existing methods model spatial and temporal features separately and then weigh them, resulting in the decoupling of spatiotemporal features.

The authors propose a spatiotemporal long short-term memory (ST-LSTM) method for fault diagnosis of rotating machinery. The authors collected vibration signals from real rolling bearing and gearing test rigs for verification.

Through these two experiments, the authors demonstrate that machine learning methods still have advantages on small-scale data sets, but our proposed method exhibits a significant advantage due to the simultaneous modeling of the time domain and space domain. These results indicate the potential of the interactive spatiotemporal modeling method for fault diagnosis of rotating machinery.

The authors propose a ST-LSTM method for fault diagnosis of rotating machinery. The authors collected vibration signals from real rolling bearing and gearing test rigs for verification.

The aim of this paper is to implement direct teaching of industrial manipulators using current sensors. The traditional way to implement teaching is either to use a teaching pedant, which is time consuming, or use force sensors, which increases system cost. To overcome these disadvantages, a novel method is explored in the paper by using current sensors installed at joints as torque observers.

The method uses current sensors installed at each joint of a manipulator as torque observers and estimates external forces from differences between joint-driven torque computed based on the values of current sensors and commanded values of motor-driven torque. The joint-driven torque is computed by cancelling out both pre-calibrated gravity and friction resistance (compensation). Also, to make the method robust, the paper presents a strategy to detect unexpected slowly drifts and zero external forces and stop the robot in those situations.

Experimental results demonstrated that compensating the joint torques using both pre-calibrated gravity and friction resistance has performance comparable to a force sensor installed on the end effector of a manipulator. It is possible to implement satisfying direct teaching without using force sensors on 7 degree of freedom manipulators with large mass and friction resistance.

The main contribution of the paper is that the authors cancel out both pre-calibrated gravity and friction resistance to improve the direct teaching using only current sensors; they develop methods to avoid unsafe situations like slow drifts. The method will benefit industrial manipulators, especially those with large mass and friction resistance, to realize flexible and reliable direct teaching.

For prefabricated building construction, improper handling of the production scheduling for prefabricated components is one of the main reasons that affect project performance, which causes overspending, schedule overdue and quality issues. Prior research on prefabricated components production schedule has shown that optimizing the flow shop scheduling problem (FSSP) is the basis for solving this issue. However, some key resources and the behavior of the participants in the context of actual prefabricated components production are not considered comprehensively.

This paper characterizes the production scheduling of the prefabricated components problem into a permutation flow shop scheduling problem (PFSSP) with multi-optimization objectives, and limitation on mold and buffers size. The lean construction principles of value-based management (VBM) and just-in-time (JIT) are incorporated into the production process of precast components. Furthermore, this paper applies biogeography-based optimization (BBO) to the production scheduling problem of prefabricated components combined with some improvement measures.

This paper focuses on two specific scenarios: production planning and production rescheduling. In the production planning stage, based on the production factor, this study establishes a multi-constrained and multi-objective prefabricated component production scheduling mathematical model and uses the improved BBO for prefabricated component production scheduling. In the production rescheduling stage, the proposed model allows real-time production plan adjustments based on uncertain events. An actual case has been used to verify the effectiveness of the proposed model and the improved BBO.

With respect to limitations, only linear weighted transformations are used for objective optimization. In regards to research implications, this paper considers the production of prefabricated components in an environment where all parties in the supply chain of prefabricated components participate to solve the production scheduling problem. In addition, this paper creatively applies the improved BBO to the production scheduling problem of prefabricated components. Compared to other algorithms, the results show that the improved BBO show optimized result.

The proposed approach helps prefabricated component manufacturers consider complex requirements which could be used to formulate a more scientific and reasonable production plan. The proposed plan could ensure the construction project schedule and balance the reasonable requirements of all parties. In addition, improving the ability of prefabricated component production enterprises to deal with uncertain events. According to actual production conditions (such as the occupation of mold resources and storage resources of completed components), prefabricated component manufacturers could adjust production plans to reduce the cost and improve the efficiency of the whole prefabricated construction project.

The value of this article is to provide details of the procedures and resource constraints from the perspective of the precast components supply chain, which is closer to the actual production process of prefabricated components. In addition, developing the production scheduling for lean production will be in line with the concept of sustainable development. The proposed lean production scheduling could establish relationships between prefabricated component factory manufacturers, transportation companies, on-site contractors and production workers to reduce the adverse effects of emergencies on the prefabricated component production process, and promote the smooth and efficient operation of construction projects.

This paper aims to propose a forcefree control algorithm that is based on a dynamic model with full torque compensation is proposed to improve the compliance and flexibility of the direct teaching of cooperative robots.

Dynamic parameters identification is performed first to obtain an accurate dynamic model. The identification process is divided into two steps to reduce the complexity of trajectory simplification, and each step contains two excitation trajectories for higher identification precision. A nonlinear friction model that considers the angular displacement and angular velocity of joints is proposed as a secondary compensation for identification. A torque compensation algorithm that is based on the Hogan impedance model is proposed, and the torque obtained by an impedance equation is regarded as the command torque, which can be adjusted. The compensatory torque, including gravity torque, inertia torque, friction torque and Coriolis torque, is added to the compensation to improve the effect of forcefree control.

The model improves the total accuracy of the dynamic model by approximately 20% after compensation. Compared with the traditional method, the results prove that the forcefree control algorithm can effectively reduce the drag force approximately 50% for direct teaching and realize a flexible and smooth drag.

The entire algorithm is verified by the laboratory-developed six degrees-of-freedom cooperative robot, and it can be applied to other robots as well.

A full torque compensation is performed after parameters identification, and a more accurate forcefree control is guaranteed. This allows the cooperative robot to be dragged more smoothly without external sensors.

The main objective of the paper is to explore the theoretical correlation of base oil viscosity in grease and to study the effect of grease grade on mechanical vibrations associated with the damaged rolling bearings.

For theoretical purposes, formulation theory of dimensional analysis was implemented. Experiments were then performed on the test bearings lubricated with three different types of greases, namely, SKF LGHP2, SKF LGMT3 and SKF LGWA2.

The numerical results obtained from the theoretical model along with the results of experiments show that the vibration amplitudes of the defective bearings come down to a lower level when it is lubricated with the grease of a higher base oil viscosity.

The promising results from the theoretical model make it usable for the practical rotating machineries applying a variety of the rolling bearings. Consequently, if the bearing is not severely damaged, its performance can be increased by lubricating it with thicker grease.

Despite many significant contributions in the field to detect the presence of defects, not many studies have been performed that relate the lubrication condition of the rolling bearings with the vibration response, because around 50-75% of the bearing failures are attributed to be lubrication related. Hence, there is need to develop a mathematical model that can correlate the vibration severity of the bearings with viscosity of the lubricant oil in the greases along with other design and operating parameters.