高级检索

基于神经网络的变电顺序控制及防误技术优化

Neural Network Based on the Transformation Sequence Control and Error Prevention Technology Optimization

  • 摘要: 为实现一次、二次设备的一体化智能防误和多重判别,研究基于神经网络的变电顺序控制及防误技术。采用具有智能防误校验的推理智能成票,通过校核层次化校核一次和二次两种设备的操作顺序以及检修票和操作票的状态等,实现一体化顺序控制;一体化智能防误优化模型采用语义分析方法和隐马尔科夫模型,完成操作票内容的分词处理和词性标注;结合语法规则获取操作票的动作、对象、状态等信息提取操作票内容后,将提取结果输入基于改进神经网络中,校正倒闸操作,实现防误技术优化。测试结果显示:该技术提取的内容完整度均在0.90以上,精准获取操作票中的内容;保证倒闸装置的动作正确程度,避免发生误动和拒动情况,实现一体化智能防误,并且实现一次、二次操作协同配合、多重判别。

     

    Abstract: In order to realize the integrated intelligent error prevention and multi-discrimination of primary and secondary equipment, the substation sequence control and error prevention technology based on neural network is studied. The reasoning intelligent ticketing with intelligent anti-error verification is adopted, and the operation sequence of the primary and secondary equipment, as well as the status of the maintenance ticket and the operation ticket, etc., are checked hierarchically to realize integrated sequence control; integrated intelligent anti-virus The error optimization model adopts semantic analysis method and hidden Markov model to complete the word segmentation and part-of-speech tagging of the operation ticket content; after obtaining the action, object, status and other information of the operation ticket in combination with the grammar rules to extract the operation ticket content, the extraction result is input based on the In the improved neural network, the switch-off operation is corrected to realize the optimization of error prevention technology. The test results show that the completeness of the content extracted by this technology is above 0.90, and the content in the operation ticket is accurately obtained; it ensures the correctness of the action of the brake device, avoids the occurrence of malfunction and refusal, and realizes integrated intelligent error prevention, and Realize the cooperation and coordination of primary and secondary operations and multiple judgments.

     

/

返回文章
返回