针对传统发电调度模式难以满足目前调度需求而出现的各种弊端,调度过程中工作票人工审核存在效率低、风险高等问题。本文基于双向长短期记忆网络(Bi-directional Long Short-Term Memory,BiLSTM)-条件随机场(Conditional Random Fields,...针对传统发电调度模式难以满足目前调度需求而出现的各种弊端,调度过程中工作票人工审核存在效率低、风险高等问题。本文基于双向长短期记忆网络(Bi-directional Long Short-Term Memory,BiLSTM)-条件随机场(Conditional Random Fields,CRF)的实体识别与行为树(Behavior Tree,BT)防误的智能防误算法的研究,提出一种智能防误调控一体化平台设计方案,给出平台功能框架,并在乌江实现了成功应用。通过对基于多种不同算法的实体识别模型的测试,验证结果表明:基于BiLSTM-CRF算法的实体识别模型的优越性;基于该智能防误调控一体化平台的研究,在乌江水电开发了乌江调控一体化平台并通过了测试,验证了本设计方案和智能防误功能的可行性和有效性。展开更多
针对中文文本检错纠错研究任务,提出了基于知识增强的自然语言表示模型(enhanced representation through knowledge integration, ERNIE)与序列标注结合的中文文本检错纠错模型。该模型由检错和纠错两部分组成,检错阶段ERNIE使用全局...针对中文文本检错纠错研究任务,提出了基于知识增强的自然语言表示模型(enhanced representation through knowledge integration, ERNIE)与序列标注结合的中文文本检错纠错模型。该模型由检错和纠错两部分组成,检错阶段ERNIE使用全局注意力机制进行词向量编码输入到BiLSTM-CRF序列标注模型中,双向长短期记忆网络(bi-directional long short-term memory, BiLSTM)提取上下文的信息进行拼接生成双向的词向量,再通过条件随机场(conditional random field, CRF)计算联合概率增加对邻近词标签的依赖性优化整个序列,从而解决标注偏置等问题给出的错误标注。纠错阶段根据检错模型输出的结果采用不同策略分类纠错,将标注为错字、缺字的错误使用ERNIE掩码语言模型和混淆集匹配进行预测,对多字、乱序错误直接纠正。实验结果表明,引入序列标注根据错误类型进行分类纠错有效提升了纠错率,在SIGHAN数据集上测试F1达到了81.8%。展开更多
In this paper, Φ-pseudo-contractive operators and Φ-accretive operators, more general than the strongly pseudo-contractive operators and strongly accretive operators, are introduced. By setting up a new inequality, ...In this paper, Φ-pseudo-contractive operators and Φ-accretive operators, more general than the strongly pseudo-contractive operators and strongly accretive operators, are introduced. By setting up a new inequality, authors proved that if ?is a uniformly continuous Φ-pseudo-contractive operator then T has unique fixed point q and the Mann iterative sequence with random errors approximates to q. As an application, the iterative solution of nonlinear equation with Φ-accretive operator is obtained. The results presented in this paper improve and generalize some corresponding results in recent literature.展开更多
文摘针对传统发电调度模式难以满足目前调度需求而出现的各种弊端,调度过程中工作票人工审核存在效率低、风险高等问题。本文基于双向长短期记忆网络(Bi-directional Long Short-Term Memory,BiLSTM)-条件随机场(Conditional Random Fields,CRF)的实体识别与行为树(Behavior Tree,BT)防误的智能防误算法的研究,提出一种智能防误调控一体化平台设计方案,给出平台功能框架,并在乌江实现了成功应用。通过对基于多种不同算法的实体识别模型的测试,验证结果表明:基于BiLSTM-CRF算法的实体识别模型的优越性;基于该智能防误调控一体化平台的研究,在乌江水电开发了乌江调控一体化平台并通过了测试,验证了本设计方案和智能防误功能的可行性和有效性。
文摘针对中文文本检错纠错研究任务,提出了基于知识增强的自然语言表示模型(enhanced representation through knowledge integration, ERNIE)与序列标注结合的中文文本检错纠错模型。该模型由检错和纠错两部分组成,检错阶段ERNIE使用全局注意力机制进行词向量编码输入到BiLSTM-CRF序列标注模型中,双向长短期记忆网络(bi-directional long short-term memory, BiLSTM)提取上下文的信息进行拼接生成双向的词向量,再通过条件随机场(conditional random field, CRF)计算联合概率增加对邻近词标签的依赖性优化整个序列,从而解决标注偏置等问题给出的错误标注。纠错阶段根据检错模型输出的结果采用不同策略分类纠错,将标注为错字、缺字的错误使用ERNIE掩码语言模型和混淆集匹配进行预测,对多字、乱序错误直接纠正。实验结果表明,引入序列标注根据错误类型进行分类纠错有效提升了纠错率,在SIGHAN数据集上测试F1达到了81.8%。
文摘In this paper, Φ-pseudo-contractive operators and Φ-accretive operators, more general than the strongly pseudo-contractive operators and strongly accretive operators, are introduced. By setting up a new inequality, authors proved that if ?is a uniformly continuous Φ-pseudo-contractive operator then T has unique fixed point q and the Mann iterative sequence with random errors approximates to q. As an application, the iterative solution of nonlinear equation with Φ-accretive operator is obtained. The results presented in this paper improve and generalize some corresponding results in recent literature.
