The Chinese Train Control Systems (CTCS) have five levels from 0 to 4 based on the levels of the European Train Control Systems (ETCS). The complicated and redundant structures ensure the reliability and safety of CTCS. However, the requirements of the next-generation train control system to optimize system structure and to reduce cost of investment and maintenance are hardly met. First, the state of the art of CTCS was summarized in this study. Second, the system structure and characteristics of the five typical control system projects, namely, NGTC, SHIFT2RAIL, Positive Train Control, European Rail Traffic Management System-Regional, and Urbalis Fluence, were analyzed. Based on the actual construction of the CTCS, the developing demand on the next-generation Chinese train control system (NGCTCS) was analyzed. The three key technologies of moving block, cognitive radio for train-to-train (CR-T2T), and combined positioning were introduced. A system scheme for the overall structure of the NGCTCS, a train-centric train control system based on combined position technology, and CR-T2T were proposed in this study. This scheme can provide reference for Chinese railways to develop NGCTCS and to adapt to the development trend of NGCTCS to ensure security, simplify structure, optimize function allocation, and reduce the system cost of construction and maintenance.

During the conformal radiotherapy, the multileaf collimator is usually used to block radiation to avoid the radiation in the normal tissue around the tumor for precise irradiation in target areas. However, there are several shortcomings in the ability of rapidity and anti-interference of precise leaf control in terms of the traditional controllers of leaf position for multileaf collimator. To ensure the rapidity of leaf position in place for short treatment time, a design method of controller of leaf position for multileaf collimator has been proposed, which can keep the leaves in place with short time as well as high accuracy. First, the motor that drives the leaf movement of multileaf collimator was taken as the controlled object, and the corresponding motor model was established. Second, the design method of the fractional order anti-windup controller of leaf position was proposed according to the controlled object. Finally, using quantum-behaved particle swarm optimization (QPSO) to ensure the accuracy of leaf position for multileaf collimator in place, the parameters of the controller of leaf position for multileaf collimator were optimized. The simulation results show that the rising and settling time of the output response of the leaf drive unit is 1.31s in the controller of leaf position for multileaf collimator using the proposed design method, which is better than the 6.4s in the proportional integral differential (PID) algorithm based on the particle swarm optimization (PSO) and the 12.55s in the fuzzy PID algorithm respectively. While the parameter tuning of the controller of leaf position for multileaf collimator is performed by the QPSO algorithm, the required iteration number is less. In addition, after adding the interference, the proposed controller can make the system recover to the much more stable operation state with a stronger ability of anti-interference. The proposed method provides great reference value for the research on the precise leaf position of multileaf collimator and the cooperation control strategy in the later stage.

No abstract