Search results

This paper aims to purpose an improved sensorless position-based force controller in gravitational direction for applications including polishing, milling and deburring.

The first issue is the external force/torque estimation at end-effector. By using motor’s current information and Moore-Penrose generalized inverse matrix, it can be derived from the external torques of every joints for nonsingular cases. The second issue is the force control strategy which is based on position-based impedance control model. Two novel improvements were made to achieve a better performance. One is combination of impedance control and explicit force control. The other one is the real-time prediction of the surface’s shape allowing the controller adaptive to arbitrary surfaces.

The result of validation experiments indicates that the estimation of external force and prediction of surface’s shape are credible, and the position-based constant contact force controller in gravitational direction is functional. The accuracy of force tracking is adequate for targeted applications such as polishing, deburring and milling.

The value of this paper lies in three aspects which are sensorless external force estimation, the combination of impedance control and explicit force control and the independence of surface shape information achieved by real-time surface prediction.

Cross operation is a common operation method in the building construction process nowadays. Due to the crossover, each other's operations are disturbed, and risks also interact. This superimposed relationship of risks is worthy of attention. The study aims to develop a model for analyzing cross-working risks. This model can quantify the correlation of various risk factors.

The concept of cross operation and the cross types involved are clarified. The risk factors were extracted from cross-operation accidents. The association rule mining (ARM) was used to analyze the results of various cross-types accidents. With the help of visualization tools, the intensity distribution and correlation path of the relationship between each factor were obtained. A complete cross-operation risk analysis model was established.

The application of ARM method proves that there are obvious risk correlation deviations in different types of cross operations. A high-frequency risk common to all cross operations is on-site safety inspection and process supervision, but the subsequent problems are different. Cutting off the high-lift risk chain timely according to the results obtained by ARM can reduce or eliminate the danger of high-frequency risk factors.

This is the first systematic analysis of cross-work risk in the construction. The study determined the priority of risk management. The results contribute to targeted cross-work control to reduce accidents caused by cross-work.