随着我国工业化程度的不断提高，对能源特别是煤炭的需求也逐渐加大，然而全国50%左右的煤矿是高瓦斯矿井，瓦斯突出是煤矿安全生产的严重威胁。瓦斯抽放是防治瓦斯突出和爆炸等煤矿事故的根本措施，进行瓦斯抽放的核心设备是液压钻机。对液压钻机进行信号采集与故障诊断，是保证液压钻机正常工作，避免生产事故与经济损失的有效手段。论文采用虚拟仪器技术、粒子群优化技术与RBF神经网络技术，对液压钻机信号采集与故障诊断系统进行了研究。 首先，完成了硬件平台的整体方案设计。在对液压钻机的工作原理与故障机理分析研究后，确定了所需采集的液压钻机特征信号，并对相应的传感器和数据采集卡进行了分析与选型，同时对信号调理与硬件抗干扰措施进行了阐述。 其次，采用虚拟仪器技术对液压钻机的信号采集软件进行了开发设计。软件包括用户界面模块、信号采集功能模块、信号处理功能模块以及数据库模块等，实现了液压钻机特征信号的采集、滤波、标度变换、实时数值显示、实时波形显示、频域分析、报警显示以及数据库访问。 再次，在硬件平台上，对液压钻机进行了传感器的安装与系统调试，并使用液压钻机信号采集软件进行了液压钻机信号采集实验。实验结果表明，硬件平台工作正常，信号采集软件各项功能达到设计要求。 最后，提出并实现了基于粒子群优化RBF神经网络算法对液压钻机的故障诊断。将粒子群优化技术与RBF神经网络技术相结合，通过分析液压钻机的故障类型，建立了故障诊断模型；采用LabVIEW与MATLAB混合编程，完成了算法与故障诊断模块的设计。与标准RBF神经网络算法的对比分析表明，粒子群优化RBF神经网络算法不仅实现了液压钻机的故障诊断，而且诊断正确率更高。 论文对液压钻机信号采集与故障诊断系统的研究，有效地提高了液压钻机工作的可靠性与安全性，同时也是对虚拟仪器技术、粒子群优化技术和RBF神经网络技术与液压钻机相结合的一次尝试，为今后进一步的研究打下了基础，具有理论意义和实用价值。

With the increasing industrialization level of China, the demand for coal energy gradually increased. However, about 50% of the whole country’s coal mines are high-gas mines, and gas outburst is a serious threat to coal mine safety production. The fundamental measure to prevent the accidents of gas outburst and explosion is gas drainage, and the core equipment is hydraulic drill rig. Signal acquisition and fault diagnosis for hydraulic drill rig is an effective method to ensure normal operation and avoid production accident. In this paper, the virtual instrument, particle swarm optimization and RBF neural network are combined to research a signal acquisition and fault diagnosis system for hydraulic drill rig. Firstly, the hardware platform is designed. Through the analysis of the hydraulic drill rig operational principle and failure mechanism, the feature signals are determined; the sensors and data acquisition card are selected. Moreover, the signal conditioning and anti-interference measures are expounded. Secondly, the hydraulic drill rig signal acquisition software is designed using virtual instrument. User interface module, signal acquisition module, signal processing module and database module are included in the software. The feature signals are acquired, filtered, scale transformation, real-time numerical displayed, real-time waveform displayed, frequency-domain analysed, alarm displayed and saved in database. Thirdly, the hydraulic drill rig feature signals acquisition experiments are done on the hardware platform using the signal acquisition software. The experimental results demonstrated that the hardware platform works properly; all functions of the signal acquisition software achieve the design requirements. Finally, the hydraulic drill rig fault diagnosis is designed through RBF neural network algorithm based on particle swarm optimization. The fault diagnosis model is designed through analysed the fault types. The algorithm and the fault diagnosis module are designed using LabVIEW and MATLAB mixed programming. In comparison with the fault diagnosis results of a standard RBF neural network algorithm, the particle swarm optimization RBF neural network algorithm not only achieves the fault diagnosis and is also higher accuracy. In this paper, the research on signal acquisition and fault diagnosis system for hydraulic drill rig can effectively improve the reliability and safety of the hydraulic drill rig. The method is not only an attempt to combine the virtual instrument, particle swarm optimization, RBF neural network and hydraulic drill rig, but also lay the foundation for further research.