近年来随着煤炭企业不断降本增效，煤炭行业减员增效势在必行，为了推进上湾智能选煤厂建设，积极探索洗选行业新的管理模式，降本增效。本文结合智能化技术发展情况，对上湾选煤厂现状及存在问题分析。针对智能化选煤厂需求，设计智能化选煤厂总体结构，重点对智能化改造方案进行设计，在现有上湾选煤厂智能化改造的基础上，通过煤块粒度识别、人员定位、数字配电、设备在线监测从而完善智能化选煤厂架构。 通过对煤块超粒度识别检测方式进行比较，确定了基于视频图像处理的大块煤识别系统，利用边缘检测算法以检测煤块横向及纵向边缘检测，最终实现煤块图像度识别，利用灰度算法，煤与矸石灰度不同，达到煤与矸石的识别。并对超粒度识别系统结构、以及基于超粒度识别技术构建的大块矸石预选系统进行说明，最后对煤块超粒度识别系统功能进行测试，测试在不同时间、矿井原煤粒度变化情况下测试，均能有效对大块煤粒度识别报警，同时能够纪录粒度曲线并进行分析。 通过对选煤厂人员定位特点进行分析，确定了基于ZigBee的智能照明人员定位系统，利用基于到达时间(TOA)、基于到达时间差(TDOA)算法结合，通过智能照明人员定位系统架构、功能设计，实现人员定位及照明控制功能，并对系统功能测试，上湾选煤厂人员定位系统可以实现选煤厂人员位置、历史轨迹、智能照明结合的选煤厂人员定位。 根据上湾选煤厂正常停送电流程，设计了选煤厂数字配电系统总体架构，搭建了系统各层次基础模型，并对各业务属性进行配置。重点按照停送电流程，对不同场景下停送电业务操作进行测试。测试在不同条件下，进行任务下达 ，该系统可以准确下达停送电任务。 通过对上湾选煤厂在线监测系统构成、数据流程及软件架构进行设计，并对监测点位置、报警值限定等进行配置，重点对不同条件下设备的振动、温度监测等功能进行测试。测试数据均能有效反应各监测点温度、振动数据。 结果表明煤块超粒度识别系统能够实时监测煤块粒度，及时发现并预警，能够有效控制煤炭产品粒度；人员定位系统能够很好的实现了人员的精确定位；数字配电系统在不同条件下都可以有效提升停送电任务效率、准确率，降低人员工作强度；在线监测系统能够很好的实现设备状态监控，做到提前预警。

In recent years, with the continuous cost reduction and efficiency increase of coal enterprises, it is imperative to reduce staff and increase efficiency in the coal industry. In order to promote the construction of Shangwan Intelligent Coal Preparation Plant, we actively explore a new management model of the Coal Preparation industry to reduce costs and increase efficiency. with the development of intelligent technology, this paper analyses the current situation and existing problems of Shangwan Coal Preparation Plant.According to the demand of intelligent coal preparation plant, the overall structure of intelligent coal preparation plant is designed. Emphasis is laid on the design of the intellectualized transformation scheme. On the basis of the existing intelligent transformation of Shangwan Coal Preparation Plant, the structure of the intelligent coal preparation plant is improved through coal particle size identification, personnel positioning, digital power distribution and on-line monitoring of equipment. By comparing the detection methods of super-granularity recognition of coal blocks, the recognition system of large coal blocks based on video image processing is determined. Edge detection algorithm is used to detect the lateral and vertical edges of coal blocks, and finally the image recognition of coal blocks is realized.Using gray algorithm, coal and gangue have different gray levels, so as to achieve the recognition of coal and gangue.The structure of super-granularity recognition system and the pre-selection system of large Gangue Based on super-granularity recognition technology are described. Finally, the function of coal block super-granularity recognition system is tested. Testing in different time and the change of raw coal particle size can effectively identify and alarm the particle size of large coal, and record and analyze the particle size curve. Based on the analysis of personnel positioning characteristics in coal preparation plant, the intelligent lighting personnel positioning system based on ZigBee is determined. By using TOA and TDOA algorithm, the personnel positioning and lighting control functions are realized through the structure and function design of the intelligent lighting personnel positioning system. The system function test and the personnel positioning system of Shangwan Coal Preparation Plant are carried out. Shangwan Intelligent Coal the personnel positioning and lighting control functions achieve the location of coal preparation plant personnel, historical track, intelligent lighting combined with the location of coal preparation plant personnel. According to the normal power cut and send process of Shangwan Coal Preparation Plant, the overall structure of the digital distribution system of the Coal Preparation Plant is designed, the basic model of each level of the system is built, and the business attributes are configured. According to the nomal process of power cut and send , the outage operation is tested under different scenarios. Testing under different conditions, the system can accurately deliver power outage tasks. Through the design of on-line monitoring system structure, data flow and software architecture of Shangwan Coal Preparation Plant, the location of monitoring points and the limit of alarm value are configurated, and the functions of vibration and temperature monitoring of equipment under different conditions are emphatically tested. The test data can effectively reflect the temperature and vibration data of each monitoring point. The results show that：the recognition system of large coal blocks can monitor the coal particle size in real time, detect and warn in time, and control the coal product particle size effectively。The personnel positioning system can achieve the accurate positioning of personnel well.The digital power distribution system can effectively improve the efficiency and accuracy of power outage tasks under different conditions, and reduce the work intensity of personnel.he online monitoring system can be very effective,Good realization of equipment condition monitoring, early warning.