Research On Position And Pose Perception Theory And Technology For Coal Mine Mobile Equipments In Chain Sensor Networks

In order to reduce coal mining accidents and mortality, digital mine needs to bevigorously developed, which can improve the information and automation level of coalmine equipment. The ultimate goal of less man or unmanned is the hotspot ininternational mining industry. Taking shearer, boring machine and transport vehicles asthe object, this dissertation research the position and pose perception theory andtechnology supported by the National High Technology Research and DevelopmentProgram of China. On the basis of chain-type sensor networks mine tunnel andmechanized mining face, the coverage and routing strategy is researched using theenergy consumption model for wireless nodes. Kernel canonical correlation analysis isapplied to find nonlinear sets of wireless signals, and the mobile target position can besolved by the constrained total least-squares. After researching the rules of wirelessdistance errors, the coordinate errors of anchor nodes, deployment density and shearerpositioning accuracy, strapdown inertial navigation system (SINS) integrated withchain-type wireless sensor networks (CWSN) is proposed for shearer position and pose.The main research work includes:1) Designing coverage and routing controlling strategy for chain-type wireless sensornetworks in long and narrow structure of coal mine monitoring. After analyzing theunbalanced consumption and hot spot of sensor nodes along chain length directions,network topology structure, including two base stations, sensor node clusters andtransmission nodes clusters are proposed. Then, setting network best lifetime asobjective function, coverage and routing controlling are researched and designed basedon the energy consumption of wireless nodes. Further, the non-uniform symmetricclusters model is proposed used in long and narrow structure of coal mine. The energybalance and link communication load performance of coal mine chain network issimulated, which can fuse the advantages of non-uniform and clustering deployments.2) Researching on position for mobile target using data correlation analysis withuncertain anchor nodes in coal mine. After analyzing the motion characteristic of coalmine mobile target, the wireless signal set correlations of similar geometric positions arediscovered in chain-type structure. However, the wireless signals presents nonlinearcharacteristic affected by sensor noises and environment noises. Kernel canonicalcorrelation analysis is applied to analyze the correlation coefficients of these nonlinearwireless signal sets. Then, two maximum correlative sets of wireless signals are integrated into a set of optimal wireless signals. Considering the coordination of anchornodes is uncertain in chain network of coal mine, constrained total least squares methodis adopted to estimate the position of mobile target, which can realize the mobile targetpositioning using the enhanced distance accuracy by data fusion.3) Establishing the accuracy evaluation model of shearer positioning using expandcramer-rao lower bound with multi-source errors. After analyzing the cooperativemovement rules of mining fleet on mechanized mining face, the machinery-tracked ofhydraulic support and adaptive height adjustment technology are researched based onshearer positioning. The chain networks and coordinate system are built on mechanizedmining face. Then, duality mapping between local strong signal set and positioningspatial domain have been established and the coordinate of anchor nodes is constrainedbased on motion characteristic of mining fleet. Further, expand cramer-rao lower boundis derived for researching inherent relationship between shearer positioning precisionand multi-source errors and coordinate error of anchor nodes.4) Proposing synchronized and reliable position and pose tracking technology forshearer integrating SINS with CWSN measurements. After researching the kinematicsand parameter constraints for mining fleet, the pose, velocity and position equations,including the corresponding error equations are derived under SINS. Due to the fact thatSINS has short-term accurate positioning performance and is known for its accumulativeerror over time, collaborative system of SINS/CWSN is adopted to estimate the shearerposition. Then, the time matching among SINS and CWSN is explored and the positionerror caused is derived. Further, multi-parameter interaction mechanism is researchedand tightly coupled model of shearer position under SINS/CWSN is established. Finally,adaptive calibration strategy for shearer position is realized under the abnormalconditions of SINS and CWSN.Research results from this dissertation play an important guiding significance to thethorough study of position and pose perception for coal mine mobile equipment and theycould provide a technical support for the widespread application of data fusion miningand multi-source cooperative sensors.