基于人工智能算法的投诉用户质差小区定位及预测的方法研究

目前LTE及GSM的质差小区投诉定位主要是依靠用户投诉的地址回放分析附近小区是否存在性能、告警上的裂变,如无则安排人员现场测试信号强度、干扰等。为了解决自动定位投诉质差小区,创新性提出以数据驱动来实现业务规则自动化,充分考虑用户投诉的时间和问题、网络交互信令、无线性能指标和告警等海量数据,运用人工智能技术中的逻辑回归及决策树算法训练出“2G语音、VoLTE语音及数据业务的投诉质差小区定界模型、质差小区预测模型及问题网元”,更快速,更智能,更准确输出语音及数据业务投诉原因、质差小区及质差原因,并对潜在质差小区进行预测,对无线规划、基站建设、故障处理、网络改造等方面的工作也具有积极指导意义。     

The impact of powder oxygen content on formability of 12CrNi2 alloy steel fabricated by laser melting deposition

Three various 12CrNi2 alloy steels were successfully fabricated on 35CrMo alloy steel by laser melting deposition with the aim of investigating the effect of powder oxygen content on formability of 12CrNi2 alloy steel. The phase constituents, macroscopic morphology, microstructure, and microhardness of powders and samples were studied by some modern analysis methods. The experimental results showed that with the increase of the oxygen content from 220 to 600?ppm, the sphericity and density of the powders gradually decreased which led to the appearance of pores in the as-deposited samples. The phase constituent of three as-deposited samples with different oxygen contents was α-Fe(M) phase. The percentage porosity in the macroscopic morphology increased with the increase of the oxygen content. Meanwhile, the high oxygen content contributed to more pores and the crack formation in the microstructure. With the increase of oxygen content, the microhardness of the as-deposited samples would decrease.     

柿竹园野鸡尾铜锡矿体锡浮选回收试验研究

柿竹园野鸡尾铜锡多金属矿含Cu 0. 18%,含Sn 0. 36%,其它有价元素均较低。由于铜锡有价元素嵌布粒度较细,磨矿粒度较细,采用重选回收率较低,因此采用浮选的方法来回收锡石。本研究采用碳酸钠作为pH值调整剂,硝酸铅作为活化剂,GYB作为捕收剂,松醇油作为起泡剂,采用一次粗选两次扫选三次精选的工艺流程,最终闭路试验结果表明,锡石精矿中锡含量为5. 27%、回收率62. 42%。     

牵引逆变器IGBT风冷散热试验及仿真平台搭建

IGBT散热性能是影响高速动车组在行驶过程中安全性、可靠性及动力性的重要因素。本文以某出口铰接车牵引变流器箱为试验研究载体,研制了专用风冷散热试验台;同时运用仿真软件Flotherm进行了仿真计算,通过实物试验与仿真计算的对比研究,验证了仿真计算的可行性。     

Context enabled decision aids to support the situation awareness and performance of risk engineers carrying out loss prevention surveys

Routine inspection by insurance companies at their clients’ facility, also known as loss prevention survey, help identify the best strategies to minimize damages when there is a high-speed wind event. More specifically, wind vulnerabilities associated with a building are evaluated using a process known as windstorm risk inspection. This routine inspection helps clients reduce the extent of damages caused by high-speed wind events including hurricane and tornado. Risk engineers make use of their subjective and analytical deduction skills to successfully carry out the inspection tasks. In this research the researchers investigated the effect of context-based visualization strategies on situation awareness and their understanding of the situation. The study examined how different types of information contribute towards the three levels of situation awareness. Following a between-subjects study design, 65 participants completed the study. Each session lasted 90–120 min. A checklist based and predictive display-based decision aids were tested and found to be effective in supporting the situation awareness requirements as well as performance of risk engineers. However, the predictive display only helped with certain tasks such as understanding the interaction among different components on the rooftop. For remaining tasks such as perceiving obvious issues like membrane tear, clogged drains and vegetation growth, checklist alone was sufficient. This study helped the understanding of the advantages and disadvantages of the decision aids tested. More specifically, these decision aids can improve the mental model of novice risk engineers. Additionally, this study provided insights that could help design training materials for infrastructure inspectors.     

Computational Fluid Dynamics Study on a Top-Blown Smelting Process with Lance Failure in an Isa Furnace

JOM - The Isa/Ausmelt smelting technology with a top submerged lance (TSL) has been extensively used in copper smelting processes. However, the TSL is extremely vulnerable to damage and failure...     

全钢载重子午线轮胎内衬层复合气泡的改善

分析全钢载重子午线轮胎内衬层复合气泡产生原因并提出相应的解决措施。通过增大复合压辊角度、硬度和压力及在气密层复合前增加位移传感器以减小气密层拉伸,使其与过渡层复合后收缩率一致,解决了内衬层复合气泡问题,提高了轮胎质量和企业经济效益。     

镁改性活性炭强化活性污泥处理混合污水研究

基于序批式活性污泥法(SBR)工艺,将镁盐改性活性炭(MgO-PAC)与传统活性炭(PAC)混合而成MPAC材料,用于处理生活与工业混合污水。通过连续30 d的运行实验,探讨了MPAC材料对生活与工业混合污水中COD、NH4^+-N和TP的去除效果以及对污泥的比耗氧速率、沉降性能和微生物多样性的影响。结果表明,投加MPAC材料对污水中COD的去除率提升了12.7百分点,对TP的去除率提升了17.5百分点,对NH4^+-N的去除率超过86.4%。投加MPAC后处理效果更好的重要原因,在于MPAC使得活性污泥的沉降性能和比耗氧速率得到明显改善,也提升了污泥的微生物丰度。MPAC对活性污泥处理生活与工业混合污水具有强化作用。     

基于蛛状网拓扑智能变电站“三层一网”通信系统研究

在110kV及以下电压等级变电站中，“三层一网”架构虽然简化了二次系统接线，但实时性和可靠性难以保证。针对传统拓扑结构下的实时性和可靠性无法同时达到最优效果的问题，本文采用“三层一网”架构下蛛状网拓扑结构，仿真结果表明该结构下的通信系统实时性最佳，解决了传统拓扑结构实时性差的问题，并结合算例验证了其可行性；运用故障树分析法定量分析不同结构的可靠性，结果表明蛛状网结构的可靠性更好。为解决研究复杂的通信系统可靠性时传统故障树方法有一定的局限性的问题，本文引入T-S模糊故障理论，建立T-S故障树模型，计算结果表明蛛状网结构下通信系统可靠性最好。根据本文研究可以得到蛛状网拓扑结构能同时提高通信系统实时性和可靠性的结论。