Formal proofs of operator identities by a single formal computation

A formal computation proving a new operator identity from known ones is, in principle, restricted by domains and codomains of linear operators involved, since not any two operators can be added or composed. Algebraically, identities can be modelled by noncommutative polynomials and such a formal computation proves that the polynomial corresponding to the new identity lies in the ideal generated by the polynomials corresponding to the known identities. In order to prove an operator identity, however, just proving membership of the polynomial in the ideal is not enough, since the ring of noncommutative polynomials ignores domains and codomains. We show that it suffices to additionally verify compatibility of this polynomial and of the generators of the ideal with the labelled quiver that encodes which polynomials can be realized as linear operators. Then, for every consistent representation of such a quiver in a linear category, there exists a computation in the category that proves the corresponding instance of the identity. Moreover, by assigning the same label to several edges of the quiver, the algebraic framework developed allows to model different versions of an operator by the same indeterminate in the noncommutative polynomials.     

基于线性规划的高温超导磁体线圈设计方法

电子真空回旋器件是一种对磁场精度要求较高的微波源装置, 一般采用超导磁体提供磁场环境. 超导磁体的应用中, 磁场分布的实现是超导磁体设计的核心问题. 提供回旋器件磁场的高温超导磁体包含较复杂的磁体绕组, 为了解决此类设计计算问题, 本文提出了一种包含设计区域约束的线性优化方法进行回旋器件高温超导绕组的设计优化, 通过分步的约束和线性优化计算, 可得到同时满足设计要求和绕组可实现的设计磁场电流分布设计. 计算实例的结果给出了一个提供磁场强度1 .3 Tesla, 长度285 mm 的均匀磁场区域, 同时满足多位置的磁场要求, 设计结果与要求一致度较好, 精度满足应用需求. 该计算方法是一种可适用于较复杂磁场要求和超导绕组结构的设计优化方法.     

基于不确定理论的铁路货运需求预测

合理预测铁路货运需求是铁路管理部门建设、运营等决策基础。为应对铁路货运需求的复杂变化,基于Pearson相关性分析方法筛选出铁路货运需求的七个具有关键影响的因素,并结合不确定理论建立不确定多元线性回归模型,相应的铁路货运预测结果由传统单一值变成可能的需求区间范围,更加符合处于不确定环境下的铁路货运需求实际情况。选取国家统计局2004~2016年相关数据进实证研究,并与回归模型以及BP模型的预测结果对比分析,实验表明不确定多元线性回归的预测结果更加精确。     

A Comparative Analysis of Machine Learning Techniques for Muon Count in UHECR Extensive Air-Showers

The main goal of this work is to adapt a Physics problem to the Machine Learning (ML) domain and to compare several techniques to solve it. The problem consists of how to perform muon count from the signal registered by particle detectors which record a mix of electromagnetic and muonic signals. Finding a good solution could be a building block on future experiments. After proposing an approach to solve the problem, the experiments show a performance comparison of some popular ML models using two different hadronic models for the test data. The results show that the problem is suitable to be solved using ML as well as how critical the feature selection stage is regarding precision and model complexity.