The invention mainly discloses a method for detecting that a storage battery car enters an elevator. The method includes the following steps that S1, the top of an elevator car is provided with a camera with voices, a target object entering the elevator car is shot, and the camera is used for acquiring data of the target object; S2, the data are decoded through a video decoder, then the data are transmitted to an algorithm plate, storage battery car detecting is performed through a deep learning algorithm, picture analyzing is performed by adopting an NPU, and analysis about whether the storage battery car exists in the target object or not is output; S3, if the storage battery car exists, an elevator control signal is generated and sent to a background server, an alarming voice is played,an elevator door cannot be closed, and the elevator stops running temporarily; and S4, when the storage battery car moves out of the elevator car, the elevator control signal is disconnected, and voice playing stops. The deep learning algorithm is adopted in the method, picture analyzing and processing are performed on the special NPU, after the storage battery car is found, voice prompting is performed in time, messaging sending and elevator door control are performed, and the situation that the storage battery car enters the elevator is found and pre-warned in time.

Method for detecting that storage battery car enters elevator

The invention discloses a supply chain information collection and data analysis method and a device based on industrial vision, and the method comprises the steps: carrying out the optical CCD sensingrecognition of an industrial supply chain based on optical collection equipment, and forming a recordable optical vision image or picture information; storing the optical visual image or picture information in a specified position of a cloud server; reading the optical visual image or picture information stored in the cloud server based on a preset time threshold to perform visual feature extraction processing, and obtaining visual feature information; analyzing and discriminating the visual feature information based on an artificial intelligence analysis algorithm, and outputting a discrimination result to a corresponding database; and when the judgment result is abnormal, generating abnormal reminding information based on the abnormal judgment result, and feeding back the abnormal reminding information. In the method of the invention, monitoring and early warning of various abnormal information of the industrial supply chain are realized.

Supply chain information acquisition and data analysis method and device based on industrial vision

The present invention is directed to methods and systems for data processing in cloud computing environments and other distributed computing environments, merging classification suitable for data classification in cloud computing environments and other distributed computing environments is disclosed. A large amount of parallelism available in a cloud computing environment is utilized; a data classification method and system of a type is provided that iteratively performs highly parallel merge classification operations that can be effectively applied to data set sizes up to the range of maximumdata sets, taking into account a large number of constraints of data storage and data retrieval operations in a cloud computing environment.

Data classification system based on computer big data

The invention discloses an intelligent monitoring system and method based on big data processing and cloud transmission, which comprises a front-end data acquisition device, a big data storage platform, a cloud computing platform and a front-end device user. The front-end data acquisition equipment transfers data to the big data storage platform, and sends out alarm after data comparison; The cloud computing platform reads the big data storage platform data for fault capture or upload fault parameter value front-end equipment users. The cloud platform adopted by the invention has an artificialintelligence learning prediction algorithm, and the prediction parameters can be adjusted in real time according to the change of the environment. Compared with an independent prediction device, thecloud platform has stronger adaptability to the new reliability.

Intelligent monitoring method and platform based on big data processing and cloud transmission

The invention discloses an elevator intelligent maintenance prediction method for predicting faults. After deep learning data are read, real-time system fault prediction of an elevator door is realized by using self-organizing learning of an RBF neural network: an input layer as the first layer assumes input observation elevator data samples, an enlarged network center can be changed, and a hiddenlayer neuron center of RBF is located in an important region of an input space through the reallocation of neural network resources and K - means clustering learning; values of eta and alpha and thevalue of iteration termination precision epsilon are given; a hidden layer as the second layer adopts a hyperbolic tangent function, system dynamics are directly involved in a network structure of a network, and a nonlinear dynamic system can be learned without knowing the system order; and information output by neurons of the hidden layer is linearly weighted and output as the output result of the whole neural network by an output layer as the third layer. The elevator intelligent maintenance prediction method makes maintenance more orderly.

Elevator intelligent maintenance prediction method and system for predicting faults

The invention discloses an elevator maintenance APP for elevator maintenance personnel and a method. Different authority modules are given to managers and common maintenance personnel; acceptance check and examination of maintenance conditions of the maintenance personnel are carried out, and working modules of the maintenance personnel are viewed; interfaces are connected to position modules of given accident elevators of which the longitudes and latitudes are confirmed by Baidu map to be correct; the independent mobile APP is informed of enabling network access and being connected to a database module of a designated cloud server; volley is used at the client to make a http request with the data format of json; and json data are analyzed at the server, so as to realize an information mutual transmission module. By means of the elevator maintenance APP for the elevator maintenance personnel and the method, the position of each elevator can be positioned in real time, and when an elevator breaks down, the maintenance personnel can be on site for maintenance in time.

Elevator maintenance APP for elevator maintenance personnel and method

The invention discloses an intelligent data acquisition method based on a small number of detachable sensors The intelligent data acquisition system comprises a gyroscope and the magnetic induction sensors The gyroscope acquires the acceleration of an elevator, the accumulated running time of the elevator, the floor induction of the elevator and whether the elevator shifts or not and deviates from the track or not, and the magnetic induction sensors acquire the door opening and closing states; after data preprocessing, the sensors perform data standardization operation according to differentdimensions; and then the data are transmitted through segmentation operation According to the intelligent data acquisition method, a small number of sensors are adopted, so that the cost is reduced,the concentration of the sensors is improved, and the sensors can be conveniently replaced by users after problems occur; the sensors are centralized in one device and are independent of the body, sothat the device is more convenient to use and saves space; and meanwhile the normal operation of the body is not influenced, and a security guarantee is provided for the body

Intelligent data acquisition system and method based on small number of detachable sensors

The invention discloses an intelligent image recognition system based on natural language understanding and image graphics, the system includes a Fourier transform module, and local eigenvector module, after comparing and analyzing the characteristics of specific characters recorded in the system, If the similarity reaches the preset value, the facial information of the person with high resolutionand multi-angles will be captured, and then the high resolution image will be divided into N blocks, and then N features will be obtained by convolution of the Gabor kernel function with each block of the image, and then these features are connected in series to form a local Gabor feature vector. LDA linear discriminant analysis module, using LDA linear discriminant analysis to obtain a global feature vector and N local feature vectors dimension reduction processing, to obtain the global classifier and local component classifier. The invention obtains the high-resolution image under the condition of successful comparison and further identifies, so as to realize accurate comparison.

Zhang Fangzhou

Papers

Intelligent monitoring method and platform based on big data processing and cloud transmission

Elevator intelligent maintenance prediction method and system for predicting faults

Elevator maintenance APP for elevator maintenance personnel and method

Intelligent data acquisition system and method based on small number of detachable sensors

Intelligent image recognition system and method based on natural language understanding and image graphics