Search results

To solve the obstacle detection problem in robot autonomous obstacle negotiation, this paper aims to propose an obstacle detection system based on elevation maps for three types of obstacles: positive obstacles, negative obstacles and trench obstacles.

The system framework includes mapping, ground segmentation, obstacle clustering and obstacle recognition. The positive obstacle detection is realized by calculating its minimum rectangle bounding boxes, which includes convex hull calculation, minimum area rectangle calculation and bounding box generation. The detection of negative obstacles and trench obstacles is implemented on the basis of information absence in the map, including obstacles discovery method and type confirmation method.

The obstacle detection system has been thoroughly tested in various environments. In the outdoor experiment, with an average speed of 22.2 ms, the system successfully detected obstacles with a 95% success rate, indicating the effectiveness of the detection algorithm. Moreover, the system’s error range for obstacle detection falls between 4% and 6.6%, meeting the necessary requirements for obstacle negotiation in the next stage.

This paper studies how to solve the obstacle detection problem when the robot obstacle negotiation.

This study aims to introduce the achievements and benefits of applying wheel/rail-force–based maintenance interval extension of the C80 series wagon in China.

Chinese wagons' existing maintenance strategy had left a certain safety margin for the characteristics of widely running range, unstable service environment and submission to transportation organization requirements. To reduce maintenance costs, China railway (CR) has attempted to extend the maintenance interval since 2020. The maintenance cycle of C80 series heavy haul wagons is extended by three months (no stable routing) or 50,000 km (regular routing). However, in the meantime, the alarming rate of the running state, a key index to reflect the severe degree of hunting stability, by the train performance detection system (TPDS) for the C80 series heavy haul wagons has increased significantly.

The present paper addresses a big data statistical way to evaluate the risk of allowing the C80 series heavy haul wagons to remain in operation longer than stipulated by the maintenance interval initial set. Through the maintenance and wayside-detector data, which is divided into three stages, the extension period (three months), the current maintenance period and the previous maintenance period, this method reveals the alarming rate of hunting was correlated with maintenance interval. The maintainability of wagons will be achieved by utilizing wagon performance degradation modeling with the state of the wheelset and the often-contact side bearing. This paper also proposes a statistical model to return to the average safety level of the previous maintenance period's baseline through correct alarming thresholds for unplanned corrective maintenance.

The paper proposes an approach to reduce safety risk due to maintenance interval extension by effective maintenance program. The results are expected to help the railway company make the optimal solution to balance safety and the economy.

In a dual-robot system, the relative position error is a superposition of errors from each mono-robot, resulting in deteriorated coordination accuracy. This study aims to enhance relative accuracy of the dual-robot system through direct compensation of relative errors. To achieve this, a novel calibration-driven transfer learning method is proposed for relative error prediction in dual-robot systems.

A novel local product of exponential (POE) model with minimal parameters is proposed for error modeling. And a two-step method is presented to identify both geometric and nongeometric parameters for the mono-robots. Using the identified parameters, two calibrated models are established and combined as one dual-robot model, generating error data between the nominal and calibrated models’ outputs. Subsequently, the calibration-driven transfer, involving pretraining a neural network with sufficient generated error data and fine-tuning with a small measured data set, is introduced, enabling knowledge transfer and thereby obtaining a high-precision relative error predictor.

Experimental validation is conducted, and the results demonstrate that the proposed method has reduced the maximum and average relative errors by 45.1% and 30.6% compared with the calibrated model, yielding the values of 0.594 mm and 0.255 mm, respectively.

First, the proposed calibration-driven transfer method innovatively adopts the calibrated model as a data generator to address the issue of real data scarcity. It achieves high-accuracy relative error prediction with only a small measured data set, significantly enhancing error compensation efficiency. Second, the proposed local POE model achieves model minimality without the need for complex redundant parameter partitioning operations, ensuring stability and robustness in parameter identification.