Alerta

储能电站的安全风险评价方法、装置及设备

Resumen de: CN122617127A

本发明提供了储能电站的安全风险评价方法、装置及设备，属于储能电站安全评价领域，其方法包括：获取储能电站的耦合指标数据构建对应的指标数据矩阵；基于指标数据矩阵中各安全风险评价指标与预设的评价指标值的差异度，确定第一权重向量；基于指标数据矩阵中各安全风险评价指标的数据离散度，确定第二权重向量；获取第一权重向量与第二权重向量对应的深层耦合特征，基于深层耦合特征的组合系数对第一权重向量与第二权重向量，得到组合权重向量；将组合权重向量输入风险评估网络，得到储能电站的安全风险等级。上述方法能够提高储能电站安全风险等级评估的准确性。

FİNANSAL ZAMAN SERİLERİ İÇİN STOKASTİK OPTİMİZASYON TABANLI YAPAY SİNİRSEL BULANIK ÇIKARIM YÖNTEMİ

Resumen de: TR2026011309A2

Bu buluş, finansal zaman serisi analizi, piyasa eğilim analizi, yatırım karar destek sistemleri, algoritmik işlem sistemleri, portföy yönetim sistemleri, finansal veri analizi sistemleri, finansal teknoloji uygulamaları, yapay zekâ tabanlı tahmin sistemleri, hesaplamalı zekâ uygulamaları, veri bilimi uygulamaları, ekonomik veri analiz sistemleri, zaman serisi tahmin sistemleri, risk analiz sistemleri ve finansal modelleme uygulamalarında kullanılabilen finansal zaman serileri için stokastik optimizasyon tabanlı yapay sinirsel bulanık çıkarım yöntemi ile ilgili olup, özelliği; sistem bileşenlerini çalıştıran bir işlemciyi (2), finansal veri kaynaklarından tarihsel fiyat ve işlem hacmi verilerini toplayan, eksik veri düzeltme, aykırı değer temizleme ve veri normalizasyonu işlemlerini gerçekleştiren veri toplama ve hazırlama modülünü (3), söz konusu verilerden RSI, MACD ve hareketli ortalama teknik göstergelerini oluşturarak model giriş değişkenlerini üreten özellik çıkarma modülünü (4), klasik öğrenme algoritmaları ile referans model eğitimi gerçekleştiren ANFIS modelinin kurulması modülünü (5), ANFIS modeline ait üyelik fonksiyonu parametrelerinin optimizasyonuna yönelik stokastik tabanlı optimizasyon parametreleri üreten stokastik optimizasyon algoritmasının geliştirilmesi modülünü (6), üretilen optimizasyon parametrelerini ANFIS modelinin üyelik fonksiyonlarına uygulayan algoritmanın ANFIS model

METHOD AND SYSTEM FOR SOLVING DISTRIBUTION NETWORK DYNAMIC SAFE SPACE FOR VIRTUAL POWER PLANT AGGREGATION

Resumen de: WO2026170655A1

The present invention relates to the technical field of safe operation of distribution networks, and disclosed are a method and system for solving a distribution network dynamic safe space for virtual power plant aggregation, the method comprising: acquiring network topology information of a distribution network, and, on the basis of the network topology information of the distribution network, performing calculation to obtain a distribution network safe region represented by hyperplanes; acquiring a resource fuzzy boundary uploaded by a virtual power plant, on the basis of the resource fuzzy boundary uploaded by the virtual power plant, performing optimization decision calculation, and using an optimization decision value to generate a fuzzy model of distribution network resources; inputting the fuzzy model of distribution network resources into the distribution network safe region represented by hyperplanes, outputting a fuzzy distribution network dynamic safe space, performing model reduction and defuzzification on the fuzzy distribution network dynamic safe space by means of an improved enhanced opposition-based search algorithm, and obtaining a distribution network dynamic safe space; and inputting the distribution network dynamic safe space into a pre-established aggregation model of the virtual power plant, and outputting a virtual power plant aggregation capacity; and, in view of an original virtual power plant aggregation capacity, calculating a virtual power plant c

INTELLIGENT IDENTIFICATION METHOD AND SYSTEM FOR PARTIAL DISCHARGE BASED ON COLLABORATIVE REASONING

Resumen de: US20260244960A1

0000 An intelligent identification method and system for partial discharge based on collaborative reasoning are provided, and belong to the field of partial discharge detection. The semantic structure of high-dimensional PD data is explicitly deconstructed through the construction of multi-scale feature information map and sparse spectrum division algorithm, and the feature subchannels strongly related to the discharge mechanism are separated, thus the semantic interference in the field data is effectively suppressed. Furthermore, the fuzzy modeler induced by channel structure is used to learn local rule sets independently, and combine with the dynamic fusion mechanism driven by cross-channel prediction consistency, so that the tolerance to voltage phase loss, background noise disturbance and equipment heterogeneity is significantly improved.

APPARATUS AND METHODS FOR GENERATING STRUCTURED DATA OUTPUTS

Resumen de: US20260244949A1

0000 Apparatus for generating structured data outputs and methods used therein include a processor and a memory connected to the processor, wherein the memory contains instructions configuring the processor to receive entity data associated with an entity, the entity data including projection data and location-based data, determine at least a selection criterion as a function of the entity data, receive from a data repository a plurality of metrics as a function of the at least a selection criterion, select at least an output parameter by applying the at least a selection criterion to a plurality of output parameters, as a function of the plurality of metrics, and synthesize, using an output generation machine-learning model trained on output generation training data, a structured data output as a function of the at least an output parameter, wherein the structured data output includes a plurality of event handler graphics.

光伏组件连接及安装结构

Resumen de: CN121461862A

The invention belongs to the technical field of solar industry, and provides a photovoltaic module connecting and mounting structure, which comprises at least two adjacent photovoltaic modules, each photovoltaic module comprises photovoltaic module glass, a corrugated back plate and a frame, the photovoltaic module glass and the corrugated back plate are bonded through a structural adhesive to form a bearing unit, and the corrugated back plate is connected with the frame. Prefabricated silica gel strips are clamped into the two sides of the bearing unit in the long edge direction, preset hole positions are formed in the frame, rivets and corner connectors are installed on the frame through the preset hole positions, and a structural welding seam is formed in the joint of the frame and the corrugated back plate through laser welding of stainless steel welding wires. A gap between the folding edge of the corrugated back plate and the photovoltaic module glass is filled with sealant, a penetrating type self-drilling screw clamping hook fixing piece is arranged between the frame bending grooves of the two adjacent photovoltaic modules, and the penetrating type self-drilling screw clamping hook fixing piece clamps and limits the lower supporting piece through a self-tapping screw.

一种基于规则与预测模型的多目标关联融合方法

Resumen de: CN122595174A

本发明涉及一种基于规则与预测模型的多目标关联融合方法，属于目标数据关联信息融合领域;构建基于知识库的先验知识规则；采用道格拉斯‑普克算法获得模型训练所需要的集合数据点，在保留几何形状的同时精简数据量，实现轨迹特征提取；构建基于时间卷积网络TCN的船舶航迹预测模型，记为TCN船舶航迹预测模型；将提取的轨迹特征作为TCN船舶航迹预测模型的输入，输出TCN网络预测结果；建立基于先验知识规则和TCN网络预测模型的目标关联方法，确定目标；本发明实现提高多目标关联融合地准确率并确保关联融合地实时性。

MEMS加速度计动态标定的模糊机器学习暂态优化方法

Resumen de: CN122596129A

本发明提供了MEMS加速度计动态标定的模糊机器学习暂态优化方法，包括：步骤1，构建MEMS加速度计响应模型；步骤2，构建记忆神经网络,设计模糊反向传播学习规则；步骤3，引入模糊规则触发时间机制进行暂态过程分析；步骤4，暂态诱发芝诺现象及暂态优化；步骤5，动态标定验证。本发明结合加速度计响应特性，引入Caputo分数阶微分模型研究系统记忆、卷积相关特性，解决传统机器学习忽略系统历史状态、暂态诱发芝诺现象、建模失真的问题，能够精准匹配MEMS加速度计动态响应过程，大幅提高动态标定精度。

一种集成多传感器反馈的猕猴桃智能催熟控制系统

Resumen de: CN122581431A

本发明公开了一种集成多传感器反馈的猕猴桃智能催熟控制系统，包括多传感器阵列、数据处理中心、催熟策略生成单元和环境调控执行装置。多传感器阵列通过分布式节点对库内堆垛果实实施多点空间采样。数据处理中心将多维度状态信息转换为果皮色泽变化率、果实硬度曲线和水分散失速率，调用改进的模糊综合评价算法进行融合分析，在内源乙烯生成速率低于有效检测阈值时自适应降维至三维评价空间，生成当前综合成熟度指数。催熟策略生成单元查询成熟度‑调控参数映射表，输出动态催熟调控参数集，由环境调控执行装置调节乙烯释放器、加热装置及加湿装置的工作状态。

基于递阶选择性关注机制的数据处理方法

Resumen de: CN122594975A

本发明涉及数据处理技术领域，涉及一种基于递阶选择性关注机制的数据处理方法，包括：获取待处理对象的表征向量；将表征向量输入递阶选择性关注分配网络，获取整体语义张量；基于整体语义张量生成分类结果，并将各级汇聚过程中生成的断点式关联因子确定为与该分类结果对应的递阶式溯因结果输出。本发明通过层级化稀疏注意力与双向反馈机制的融合，在提升分类精度的同时，实现了保真、固有的、可进行层级归因追溯分析的层级化特征归因，无需依赖外部解释器。

一种基于模糊规则改进的茶叶拼配方案动态调控方法

Resumen de: CN122596455A

本发明公开了一种基于模糊规则改进的茶叶拼配方案动态调控方法，包括以下步骤：加载茶叶拼配配方库，从茶叶拼配配方库获取目标茶叶拼配方案的配方、茶叶品质和物理兼容性；获取目标茶叶拼配方案的茶叶成本；茶叶成本、茶叶品质和物理兼容性，构成目标茶叶拼配方案的属性集合；加载拼配动态调控模型，向拼配动态调控模型输入属性集合，获取目标茶叶拼配方案的权重组；目标茶叶拼配方案的权重组由品质、成本、稳定性的权重变量w、c、s构成；基于权重组计算目标茶叶拼配方案的得分，实现茶叶拼配方案动态调控。根据上述技术方案，可以实现多目标间的智能动态权衡，提升拼配方案寻优精准性；实现方案可行性的精细化，提升实用性。

一种铁路动环多源数据融合智能运维管控方法及系统

Resumen de: CN122594781A

本发明具体为一种铁路动环多源数据融合智能运维管控方法及系统，涉及铁路机房动力环境监控技术领域，包括：多源数据采集预处理模块；异构数据时空对齐模块；多维特征冲突消解模块；运行状态评估预警模块；智能联动执行管控模块。本发明中，基于奇异值分解的自适应降噪方法，能够较好地滤除各类随机噪声和干扰，同时采用高阶插值方式进行数据补全，保障了时间序列的完整性和连续性；构建了考虑环境因素影响的三维空间距离矩阵，反映不同传感器之间的物理空间关系，结合三次样条升采样与互相关相位补偿技术，实现了多源异构数据在时间和空间维度上的对齐。

一种面向机器人仿真器的模糊测试方法和系统

Resumen de: CN122594161A

本发明公开了一种面向机器人仿真器的模糊测试方法和系统，涉及模糊测试技术领域，包括：确定各种子程序的语义阶段，将语义阶段关联至种子程序形成目标种子程序确定目标种子池；以目标种子程序为状态空间，以多层级变异操作符为动作空间，通过多臂老虎机算法基于测试反馈奖励输出待变异目标种子程序的最佳操作符；结合机器人仿真器的代码文档以及在各语义阶段的可变异对象，按照最佳操作符生成变异测试用例；通过机器人仿真器执行变异测试用例确定测试反馈指标；基于测试反馈指标更新最佳操作符的测试反馈奖励和目标种子池，并通过多臂老虎机算法输出新采样的待变异目标种子程序的最佳操作符，直至测试结束。基于上述方案，提升了测试覆盖率。

一种基于模糊-AI虚拟同步机的构网型储能系统控制方法及装置

Resumen de: CN122600074A

本发明涉及储能调频控制技术领域，公开了一种基于模糊‑AI虚拟同步机的构网型储能系统控制方法及装置，控制底层构建基于经典比例‑微分控制理论的基准模型；在基准层之上引入由模糊逻辑控制器与径向基函数神经网络构成的双层智能补偿机制；模糊逻辑控制器对系统宏观运行状态进行非线性增益调度获取模糊逻辑补偿量；RBF神经网络对系统微观动态特性进行精细调节获取RBF神经网络补偿量；最终将三者协同合成作为虚拟同步发电机算法内核的虚拟参数，生成最优控制指令。与现有技术相比，本发明通过硬件模块的优化配置与软件控制策略的协同设计，结合模糊‑AI自适应控制算法，实现虚拟惯量与阻尼系数的动态自适应优化，有效提升频率稳定性、抑制功率振荡。

FCM聚类约束ERT与AMT联合反演方法及系统

Resumen de: CN122592506A

本申请提供了一种FCM聚类约束ERT与AMT联合反演方法及系统，涉及地球物理勘探技术领域。方法包括：获取隧道勘察区域的ERT与AMT同步观测数据并进行预处理；建立非结构化网格模型，初始化电阻率模型及FCM聚类参数；以非结构化网格为载体构建综合目标函数；采用交替迭代策略求解目标函数，通过固定聚类特征更新反演模型、固定反演模型更新聚类特征的交替优化至收敛获取联合反演结果。本发明通过引入模糊C均值聚类约束与交叉梯度约束的双重结构耦合机制，提升了断层破碎带及富水构造等不良地质体边界的识别精度，实现了两种物探数据的深度融合与优势互补，为隧道勘察提供了高可靠性的电阻率模型及量化表征反演不确定性的隶属度场。

一种基于机器学习的工序过程质量波动源识别方法及装置

Resumen de: CN122596723A

本发明公开了一种基于机器学习的工序过程质量波动源识别方法及装置，所述方法包括：获取工序或零件的质量特性测量数据，建立异常模式矩阵；区分异常原因是偶然因素或系统因素；若异常原因是系统因素，则进行工序过程质量波动源识别，建立模糊关系矩阵；构造异常出现可能性矩阵，对其中各元素进行归并处理，根据可信度值大小进行异常原因诊断；对生产过程中“人机料法环”各分项的潜在质量风险进行工序风险指标设定并计算工序的潜在质量风险系数，从而确定工序总风险系数，用以对工序风险进行评估；读取工序质量BOM中的工序质量因素变更信息，通过数据集成方式，输出显示工序过程质量波动源识别结果集。

基于大语言模型与模糊层次分析法的城市消防装备配置方法与系统

Resumen de: CN122596479A

本发明公开了一种基于大语言模型与模糊层级分析法的城市消防装备配置方法，方法包括：离线结构先验建模，基于大语言模型从领域标准文档中提取两层评估层次结构与方案层装备‑准则关联权重矩阵；上下文感知与多智能体模糊判断生成，通过在线检索与多智能体并行评估，得到模糊判断矩阵集合；基于认知不确定性的多智能体置信度加权聚合，将多智能体输出聚合为单一模糊判断矩阵；一致性校正与模糊重构，将单一模糊判断矩阵校正为一致性合法模糊判断矩阵；全局权重合成与增量式资源映射，由一致性合法模糊判断矩阵合成全局准则权重，并最终输出物理装备增量与最终配置。

融合机器视觉与深度学习的织物缺陷实时检测方法及系统

Nº publicación: CN122597277A 18/08/2026

Solicitante:

南通海汇科技发展有限公司

Resumen de: CN122597277A

本发明涉及织物缺陷检测技术领域，尤其涉及一种融合机器视觉与深度学习的织物缺陷实时检测方法。该方法包括：采集并配准织物的可见光与近红外波段图像，通过计算光谱响应差异并融合构建光谱融合特征；采用两级检测策略，先全局扫描定位异常区域，再自适应调整感受野修正缺陷边界；最后提取缺陷区域的几何与灰度属性，匹配分类规则库确定缺陷类别。本发明实现了对织物缺陷的高精度、自适应实时检测与分类。