13results about How to "Realize automatic diagnosis" patented technology

The invention discloses a thyroid nodule automatic diagnosis method based on multi-modal deep learning. After ultrasonic image data, CT image data and clinical data of a patient are collected and preprocessed, spatial alignment of multi-modal data is achieved; extracting deep features from the multi-modal image data by using a convolutional neural network, performing image segmentation on a thyroid nodule region by using an improved ResNet structure and a U-Net structure, and extracting a region of interest; feature extraction and fusion are realized; constructing a deep neural network of a mixed mode, and training and optimizing a deep learning model; the model reasoning comprises the steps of preprocessing new input data, extracting and fusing features and generating a final diagnosis result. According to the invention, the automatic diagnosis of the thyroid nodules is realized by fusing the multi-modal data from different image sources and combining with the advanced deep learning model, so that the accuracy and efficiency of diagnosis are improved.

The present invention discloses a method for diagnosing power failure of a nuclear power unit and assisting decision-making, comprising the following steps: step S1, combing the upstream and downstream data of electrical nodes according to the electrical primary diagram of the nuclear power unit, establishing a structured data table of electrical primary equipment and storing it in a graph database; step S2, analyzing the response of each electrical equipment under different power failure fault types based on the working principle of the electrical equipment of the nuclear power unit, establishing a knowledge base for power failure analysis of electrical equipment; configuring auxiliary decision-making schemes for different power failure fault types based on technical documents related to power failure fault processing, and establishing an auxiliary decision-making knowledge graph; step S3, establishing an auxiliary decision-making model based on a digitized electrical primary diagram, the power failure analysis knowledge base and the auxiliary decision-making knowledge graph; step S4, obtaining the power failure operating state parameters when the nuclear power unit fails and inputting them into the auxiliary decision-making model, and obtaining the fault type and the corresponding auxiliary decision. The present invention can make power failure auxiliary decisions in a timely manner.

The invention relates to a mining dump truck lifting system with a load capacity detection function and a using method thereof, the mining dump truck lifting system comprises a lifting pump connected with a hydraulic oil tank, the lifting pump is connected with a lifting oil cylinder through a lifting proportional valve, a rodless cavity of the lifting oil cylinder is provided with a pressure sensor, a lifting angle sensor is arranged on a carriage of a mining dump truck, and a slope sensor is arranged on a truck body. The system further comprises a controller, the input end of the controller is connected with the pressure sensor, the lifting angle sensor and the gradient sensor, and the output end of the controller is connected with the lifting pump and the lifting proportional valve. The controller controls the lifting proportional valve to drive the lifting pump to lift in a small range, pressure value calibration in the reference state, the no-load state and the full-load state is achieved, after calibration of the three states is completed, load capacity detection is conducted, the load capacity of the vehicle is calculated, and the system is few in part, easy to install and maintain, low in cost and high in reliability. Automatic diagnosis of the detection system can be realized, so that system errors are reduced, and faults are eliminated.

The invention discloses an automatic cleaning and maintenance control method and system for water quality monitoring equipment and a medium, and belongs to the field of water quality monitoring. The method comprises the steps that pure water generated by a built-in pure water production assembly of equipment serves as a measurement anchor point, and whether sensors are normal or not is judged on the basis of comparison results of detection numerical values of the sensors in a pure water environment and a preset pure water threshold value; monitoring operation state parameters of the pure water production assembly, and dynamically adjusting and updating the pure water threshold value corresponding to each sensor; when the value of the sensor deviates from the threshold value, first-stage pure water flushing is executed; if not, second-stage maintenance is started according to the sensor type, the optical sensor performs optical path self-cleaning or backwashing, and the electrochemical sensor performs electrode cleaning; and detecting again after maintenance, and locking the sensor and generating a fault alarm if the sensor is still not recovered. Through dynamic threshold configuration, progressive maintenance and sensor adaptive cleaning, the problems of reference drift and deep pollution are solved, the data credibility is improved, and the operation and maintenance cost is reduced.

The present invention discloses a method and device for diagnosing electric meter faults using a stacked recurrent neural network, the method comprising the following steps: obtaining current time series data associated with the electric meter to be diagnosed, inputting the preprocessed current time series data into a trained stacked recurrent neural network model, and obtaining an output corresponding to the current time series data, wherein the output is an L-dimensional vector, and L is equal to the total number of fault categories that the model needs to distinguish; and determining the diagnostic category of the electric meter to be diagnosed according to the output corresponding to the current time series data. The present invention can effectively characterize different types of electric meter faults and realize automatic diagnosis of electric meter faults; moreover, the input of the stacked recurrent neural network model is current time series data, which can effectively extract dynamic time series features rather than independent static features, thereby improving the model's characterization capability and versatility for fault types.

The present application discloses a gynecological tumor image processing system and method thereof, which uses imaging technology to collect gynecological tumor images, and introduces image processing and analysis algorithms at the back end to automatically analyze gynecological tumor images to determine the type of gynecological tumor. In this way, automatic diagnosis and suggestion output of gynecological tumor images can be achieved, providing auxiliary decision-making for doctors.

The invention discloses a fan rotating speed control method and system, and relates to the technical field of fan control, and the fan rotating speed control method comprises the steps that the rated rotating speed, the rated current, the rated voltage, the real-time rotating speed, the real-time current and the real-time voltage of a fan are obtained, and a fan operation deviation coefficient is calculated; the fan output deviation coefficient is calculated according to the real-time output wind speed, the real-time output wind direction and the real-time output wind pressure of the fan, the fan adjustment deviation coefficient is calculated according to the real-time temperature, the real-time air pressure and the real-time humidity of the fan adjustment area, and then the fan control deviation index is calculated according to the fan output deviation coefficient and the fan adjustment deviation coefficient; and obtaining fan operation deviation coefficients and fan control deviation indexes of all the fans, calculating a rank correlation coefficient, and importing the fan operation deviation coefficients, the fan control deviation indexes and the rank correlation coefficient of all the fans into a machine learning model to build a control adjustment model, so that the control adjustment model determines an adjustment value of the operation deviation coefficients. And remote monitoring and intelligent adjustment of the fan can be realized.

Embodiments of the present application provide a kind of fault detection method, device and equipment of mutual inductor and mutual inductor, it is related to power system technical field.The method comprises: obtaining the multiple sensor data that multiple sensors deployed in mutual inductor are collected in real time;According to the first feature information corresponding to multiple sensor data, determine the fault score of mutual inductor;Obtain the historical sensor data of mutual inductor, and according to multiple sensor data, historical sensor data and the equipment related information of mutual inductor, determine the dynamic fault threshold corresponding to mutual inductor;If according to fault score and dynamic fault threshold, it is determined that mutual inductor exists fault, then according to the second feature information corresponding to multiple sensor data, determine the fault type of mutual inductor.The method of the present application can comprehensively, accurately and efficiently realize the fault detection of mutual inductor, to help guarantee the stable operation of power system.

This invention discloses a method and system for video diagnosis of power system faults based on spatiotemporal semantic fusion, relating to the fields of power system fault detection, computer vision, and multimodal information fusion. The method involves acquiring fault videos and dividing them into multiple video sub-segments. For each video sub-segment, a CLIP image encoder is used to extract spatial feature vectors from each frame, resulting in a corresponding spatial feature sequence. This spatial feature sequence, after being embedded with positional information, is input into a temporal Transformer encoder. A self-attention mechanism is used to calculate the internal temporal relationships, and these are aggregated to obtain a sub-segment feature vector representing the dynamic features of that video sub-segment. Text describing the power system fault is collected, and a CLIP text encoder is used to extract text feature vectors for each fault text. Cross-modal similarity matching is performed between the sub-segment feature vectors and the text feature vectors, and the matching result with the highest similarity score is used as the fault diagnosis result. This invention has significant advantages for the identification of complex dynamic fault events.

The invention discloses a photovoltaic power station assembly hot spot fault automatic diagnosis method and system, and relates to the field of photovoltaic assembly hot spot fault diagnosis, and the method comprises the steps: collecting local environment data, power output data and assembly temperature data of a target photovoltaic power station; based on the environmental data and the power output data, obtaining a relationship between the environmental factors and power output; acquiring the output power of the photovoltaic modules in different plots under different dust accumulation conditions, and obtaining the relationship between the different dust accumulation conditions and the power output; and judging the level of the hot spot fault of the photovoltaic power station module based on the characteristic of power loss caused by the module hot spot and the module temperature data. According to the invention, the solar radiation data and the photovoltaic module temperature data are collected, and the detection and recognition algorithm of the local shadow problem is combined, so that the automatic diagnosis of the hot spot fault of the photovoltaic module and the determination of the optimal cleaning period of the photovoltaic module are realized, and the power generation efficiency and the yield of a photovoltaic power station are effectively improved.

The invention discloses an automatic diagnosis method, system and device for temporomandibular joint disorder, a medium and a product, and relates to the technical field of medical treatment.The method comprises the steps that medical history information and clinical examination record sheet information of a patient are obtained; according to the medical history information and the clinical examination record table information, whether the patient suffers from the temporomandibular joint disorder disease or not is diagnosed, and a diagnosis report is generated. The TMD patient can be automatically diagnosed, a reference is provided for a doctor, and the problem that oral treatment is affected due to the temporal-mandibular joint problem is solved.

The present invention belongs to the technical field of touch display screen management and control, and specifically is a touch display screen fault warning system based on Internet of Things data analysis, comprising a server, a touch performance detection and evaluation module, a display performance detection and evaluation module, an operation abnormality auxiliary comprehensive inspection module and a fault warning terminal; the present invention detects and analyzes the touch performance of the touch display screen through the touch performance detection and evaluation module, detects and analyzes the display performance of the touch display screen through the display performance detection and evaluation module, and the operation abnormality auxiliary comprehensive inspection module performs auxiliary monitoring on the operation abnormality of the touch display screen and generates an operation auxiliary test unqualified signal or an operation auxiliary test qualified signal, thereby realizing automatic diagnosis of touch display screen operation faults, having a high degree of intelligence, more comprehensive fault diagnosis analysis, more accurate fault diagnosis results, effectively ensuring the safe and stable operation of the touch display screen, and significantly improving the service life of the touch display screen.

The present invention relates to the technical field of ophthalmic disease diagnosis, and discloses an intelligent diagnosis and treatment management method and system for ophthalmic diseases based on a large model, the method comprising: obtaining a fundus image of a patient to be diagnosed; inputting the fundus image of the patient to be diagnosed into a fundus disease classification and segmentation model to obtain the ophthalmic disease category and disease marker segmentation result corresponding to the fundus image to be diagnosed; inputting the ophthalmic disease category, disease marker segmentation result and clinical records of the patient to be diagnosed into a large language model to obtain a fundus disease treatment and management plan for the patient to be diagnosed; the large language model is based on the LLaMA model, and a LoRA layer is added to each multi-head attention layer of the LLaMA model; an Adapter layer is added between each feedforward layer and the normalization layer of the LLaMA model. The present invention can realize the rapid diagnosis and treatment of ophthalmic diseases, improve the diagnostic efficiency, and reduce the medical burden of patients.