415results about How to "Reduce labor maintenance costs" patented technology

The invention relates to a resource allocation system in a private cloud environment. The system comprises a global policy controller, cloud application policy controllers and group policy controllers, wherein a resource manager interacts with a resource exploring proxy, receives index data, judges whether an alarm is generated for the index data or not according to alarm conditions, and if so, submits the alarm to a policy control network; the group policy controller first processes the alarm, generates a resource adjusting plan if the alarm can be processed, and reports the alarm to the cloud application policy controller if the alarm cannot be processed; the cloud application policy controller processes the received alarm, generates the resource adjusting plan if the alarm can be processed, and reports the alarm to the global policy controller if the alarm cannot be processed; the global policy controller processes the received alarm and generates the resource adjusting plan; the policy control network returns the resource adjusting plan to the resource manager; and the resource manager executes the resource adjusting plan. By constructing graded control policies and assisting coordination mechanisms among the control policies in different grades, the invention effectively improves the utilization rate of private cloud resource.

The invention discloses a machine reading model training method and device, a question and answer method and a device, and belongs to the field of natural language processing. The machine reading model training method comprises the following steps: acquiring a training sample, wherein the training sample comprises a sample question and a corresponding sample article, a real initial position and areal final position of the corresponding answer in the sample article; extracting a question feature vector of the sample question and an article feature vector of the sample article, and using a neural network structure to fuse and process the question feature vector and the article feature vector to form a fusion result; inputting the fusion result into a classifier to perform prediction of theinitial position and the final position of the answer; performing error calculation on the predicted initial position and final position, and the real initial position and the real final position of the answer; and optimizing the neural network structure according to the error calculation result. According to the machine reading model training method and device, the question and answer method anddevice in the embodiment of the invention, the corresponding answer can be directly extracted from the entire associated article through end-to-end deep learning.

The invention relates to a device and a process for preparing struvite crystals from nitrogen and phosphorus in sewage. The device comprises a reactor (5), a pH (potential of Hydrogen) controller (25), a pH regulating pump (24), an alkali liquor tank (23), a nitrogen and phosphorus containing waste water tank (22) and a secondary sedimentation tank (19), wherein the reactor (5) is composed of a reactor feed pipe (9), a reactor crystal collecting pipe (8), a reactor growth tube (3) and a reactor sediment tube (2) from bottom to top. The process comprises the steps of pumping nitrogen and phosphorus containing wastewater into the reactor (5) and pumping a magnesium chloride solution into the reactor (5) to ensure that the molar ratio of NH3-N to PO43-P to Mg <2+> in the reactor (5) is (1:1:1)-(1:1:1.3); controlling the reflux ratio between 14 and 28 and the pH equal to 6.4-9.2; and forming the struvite crystals in the reactor (5) by using the nitrogen, phosphorus and magnesium containing solution. The device disclosed by the invention has the advantages of simple structure and small floor space, is simple and easy to operate and difficult to scale, can be widely applied to preparation of the struvite crystals from the nitrogen and phosphorus containing wastewater such as coking wastewater, pig manure wastewater or sludge fermentation liquor, and is capable of realizing the resource recycling of nitrogen and phosphorus.

The invention discloses a company announcement processing method for multi-task learning and a server, and the method comprises the steps: inputting historical announcement data into a sharing layer of a multi-task learning model, and carrying out the pre-training of the historical announcement data through Bert; inputting the data set corresponding to the processing task into a task layer of themulti-task learning model to train the multi-task learning model; obtaining current announcement data, and inputting the current announcement data into the trained multi-task learning model to obtaina task processing result. The multi-task learning model is constructed through using in a transfer learning and multi-task learning mode, and the method has the advantages of being high in learning efficiency, high in generalization, low in manual maintenance cost, high in accuracy of multiple tasks, high in recall rate and convenient for engineering deployment and maintenance.

The embodiment of the invention provides a handlingmethod for switching driving modes of a rail transit train. The method comprises the steps of acquiring status information of the train after the train enters a common management area, wherein the common managementarea is an overlapping area of different operation systems of the train; obtaining a second driving mode of a to-be-switched operationsystem corresponding to a first driving mode of a current operation system of the train according to the status information of the train and preset triggering conditions; if the confirmation action ofthe second driving mode of the to-be-switched operation system is received, switching the first driving mode to the second driving mode. The handling method for switching the driving modes of the rail transit train acquires the second driving mode corresponding to the first driving mode of the train through the status information of the train and the preset triggering conditions, and achieves seamless switching between the driving modes of the train operation systems, so that manpower maintenance costs are reduced, and management efficiency of train operation is improved.

The invention discloses a multi-terminal smart home control platform based on cloud computation and raspberry Pi. Due to popularization of a network and gradual development of a home smart system currently, the home smart system obtains more and more attention. The invention develops a set of multi-terminal smart home control platform based on cloud computation and raspberry Pi. According to the method, mainly a special virtual server is used as control core basic erection of a smart home control platform, and by setting up a virtual platform in the special virtual server and configuring related services in the virtual platform to provide data services for the integral set of smart home control platform, collection, processing, use and management on control parameters of related bottom equipment of the platform are implemented, all data stream generated by the platform is transmitted in a virtual network, and an encryption and decryption protocol is set at a data receiving position ofeach port so as to ensure data safety of the system.

In the invention, the transmission of encrypted multicast data is realized only after once encapsulation, by configuring the ACL rule of multicast data stream. By the invention, the time of processing multicast data by small node can be reduced.

The embodiment of the invention discloses a processing method and device in a question-answering system based on a knowledge graph. The method comprises: after a question sentence is received, obtaining a main entity corresponding to the question sentence in a preset knowledge base; selecting at least two candidate main entities from the main entities; taking each candidate main entity as a root node, searching an adjacent edge and an adjacent node of the root node in a pre-stored knowledge graph, searching an adjacent edge and an adjacent node of the adjacent node in the next layer through the adjacent node, and so on, until a node of the last layer is searched, and obtaining a path corresponding to each candidate main entity; calculating the similarity between the text information corresponding to the path of each candidate main entity and the text information of the question sentence; selecting a path of which the similarity meets a preset highest similarity judgment condition as acandidate path of a candidate main entity; obtaining a final selection path of the main entity from the candidate paths of the candidate main entity; and determining text information corresponding tothe final selection path as an answer to the question.

The invention relates to an ocean wave energy and solar energy integrated power generation device, which comprises a concave base fixed on a coastline, and a solar energy battery acquisition part arranged on the concave base. A water flow channel is arranged on the lower part of the concave base in a penetrating manner; a baffle driven by sea waves is arranged in the water flow channel; a hydraulic rod and a hydraulic cylinder both being driven along the flowing direction of the sea waves are arranged on the concave base; a rotating shaft is fixedly arranged on the middle part of the baffle; the baffle is arranged in the water flow channel through the rotating shaft; the lower part of the baffle does reciprocal pendulum motion under the driving of the sea waves; a rotating shaft support point enables the upper part of the baffle to drive the hydraulic rod and the hydraulic cylinder to do reciprocating motion for power generation.

The invention provides a network attack path prediction method. Because network asset information is monitored in real time throughout the day through active scanning and passive scanning, the latest news of assets can be known in real time. According to the latest network asset information, the connection condition of each asset is identified, and an asset connection information and network asset topology map is drawn. Then, all possible paths of network attack are analyzed and judged, and the network administrator takes measures in advance to defend the network system and the core assets according to all the possible network attack paths predicted. Therefore, through the network attack path prediction method of the invention, network attacks can be prevented, the core data of enterprises can be protected from being invaded and stolen, and the cost of human maintenance can be reduced.

The invention provides a building electromechanical system fault intelligent prediction method and system, and the method comprises the steps: step 1, building a BIM model of a building; step 2, adopting the technology of the Internet of Things to monitor important electromechanical equipment in real time, dynamically collecting monitoring data, wherein the important electromechanical equipment iscentral and subarea control and power equipment of an electromechanical system; step 3, collecting repair work order information, performing semantic recognition on each repair work order, and matching with the BIM information to determine a repair space and an associated electromechanical system or device; step 4, establishing an electromechanical equipment fault prediction model by adopting principal component analysis and a neural network algorithm, and performing machine learning; step 5, using a cross validation method for network training until the accuracy of the obtained artificial neuron network model is available; and step 6, applying the artificial neuron network model to perform fault prediction, and notifying maintenance personnel of potential faults to perform key inspection. According to the scheme of the invention, the method can achieve the accurate prediction of the faults of the building electromechanical equipment, reduce the sudden faults of the building electromechanical equipment by 20%, guarantee the stable operation of a large public building, and reduce operation and maintenance cost.

The invention relates to a full-automatic screw length automatic detecting and sorting apparatus, which comprises an automatic feeding mechanism, a conveying guide rail is arranged in the automatic feeding mechanism, a screw moves along the conveying guide rail in a vertical state, the movement direction of the conveying guide rail is set as X direction, and the direction which is horizontal and vertical to the X direction is Y direction; the end part of the conveying guide rail is a screw length detection station, a compaction cylinder is arranged over the screw length detection station; a detection cylinder is arranged below the screw length detection station; a material taking cylinder is arranged over the station to be detected along the Y direction; a pair of screening cylinders are arranged over a screening station, a sliding pull block is arranged between the screening cylinders, a qualified product channel and an unqualified product channel are respectively arranged at two sides of the screening station along the X direction, the qualified product channel is communicated to a qualified product box, and the unqualified product channel is communicated to an unqualified product box; and a displacement sensor, the compaction cylinder, the detection cylinder, the material taking cylinder and the screening cylinder are connected to a control center signal of the screw length automatic detecting and sorting apparatus.

The invention relates to an automatic condensing fin cleaning device and an air conditioner with the cleaning device. The cleaning device is arranged on an air conditioner outer machine box body. The air conditioner outer machine box body is provided with a fin heat exchanger and a draught fan and comprises a liquid storage device, an electromagnetic valve, a spray nozzle device and a control box. The liquid storage device is arranged on the air conditioner outer machine box body and connected with the electromagnetic valve through a pipe. The electromagnetic valve is connected with the spray nozzle device through a pipe. The spray nozzle device is arranged on the fin heat exchanger. The electromagnetic valve is connected with the control box through a circuit. The automatic condensing fin cleaning device and the air conditioner with the cleaning device have the advantages that cleaning liquid is injected into the liquid storage device, the control box controls the electromagnetic valve to be opened, and the cleaning liquid in the liquid storage device is evenly sprayed onto the fin heat exchanger through the spray nozzle device, so that the purpose of cleaning fins is achieved; the problem that time and labor are consumed in the manual cleaning process is solved, automatic cleaning of the fins of a condenser is achieved, and the manual maintenance cost of a unit is lowered.

The invention discloses a geological disaster monitoring system, which comprises a monitoring station point, a platform server and a user terminal. The monitoring station uploads the monitoring data to the platform server. The platform server carries out centralized calculation on the monitoring data uploaded by each monitoring station point, then carries out filtering statistics and generates theearly-warning information. Each user terminal receives the calculation data and the early-warning information uploaded by the platform server and corresponds to the monitoring station point. The invention further discloses a geological disaster monitoring method. According to the invention, the geological disaster monitoring items of various places can be managed and resolved in a unified manner.Therefore, the working efficiency of a technical service provider is improved, and the labor maintenance cost thereof is reduced. Meanwhile, the equipment cost and the labor cost of the user side arereduced. The system and the method provide more effective, more economical and more convenient geological disaster monitoring services for all parties.

The invention discloses a remote water supply control system which is capable of achieving real-time fault monitoring early warning and sending fault warning short messages and high in degree of intelligence. The remote water supply control system comprises at least one water supply equipment end and a GPRS module which is connected with the water supply equipment end. The water supply equipment end is connected to a network server through the GPRS module, and the network server is further connected with a control center. The remote water supply control system is high in degree of intelligence, wired constraint can also be removed, the fixed IP application and cost can be saved, the control center can further remotely monitor the working condition of water supply equipment at the water supply equipment end in real time and conduct real-time monitoring early warning on faults, the fault warning short messages are sent to specified mobile phone users, so that the modernization development needs are met, and high promotional value is achieved.

An embodiment of the invention discloses an unmanned aerial vehicle power supply system and a maintenance method thereof, and an unmanned aerial vehicle charging method, and relates to the technical field of unmanned aerial vehicles, which can be used to monitor the unmanned aerial vehicle power supply system remotely so as to achieve automatic charging of the unmanned aerial vehicle in a remote area. The unmanned aerial vehicle power supply system includes a plurality of photovoltaic power supply devices, and a remote control platform connected to each of the photovoltaic power supply devices; each of the photovoltaic power supply devices includes an unmanned aerial vehicle apron including a photovoltaic battery, and a positioning communicator which is signally connected to the remote control platform and the unmanned aerial vehicle, wherein the positioning communicator is used to determine the location information of the unmanned aerial vehicle apron, and send the location information to the corresponding remote control platform and/or the corresponding unmanned aerial vehicle according to a positioning request of the remote control platform and / or the unmanned aerial vehicle.The unmanned aerial vehicle power supply system is used for charging the unmanned aerial vehicle.

The invention relates to a lighting management control system, and belongs to the field of lighting lamps. The lighting management control system is composed of a lighting managing center, an Internet center, an integrated manager and a single-lamp controller. The lighting managing center is a controlling core. The Internet center receives control demands and various data from the integrated manager and transmits the control demands and the various data to the integrated manager. The integrated manager is a communication bridge. The demands and the data are received from the lighting managing center and transmitted to the lighting managing center in a Gprs mode, and at the same time, the demands and the data are received and transmitted to the single-lamp controller in a Zigbee mode. The single-lamp controller controls different types of lighting light sources through a Zigbee wireless ad-hoc network and enables various running data of a lighting product to be automatically transmitted to the integrated controller. The lighting management control system can conduct remote wireless control, is large in control amount, high in energy conservation efficiency, convenient to control, simple to install, and safe and reliable due to the fact that the different types of light sources are used.

The embodiment of the invention provides a data storing and reading method and device using protobuf. The method includes the steps of obtaining to-be-stored data, establishing a target information example used for storing the to-be-stored data, defining the name of the target information example according to preset rules, obtaining target correlation configuration information corresponding to the target storage identification related to the name of the target information example from a proto package file, and storing the target information example in the storage position indicated by the target correlation configuration information in a database, wherein the storage identification related to the name of the information example and the correlation configuration information corresponding to the storage identification are stored in the proto package file. Compared with the prior art, no configuration files are needed, consistency of proto source files and configuration files does not need to be manually maintained, and the manual maintenance cost is greatly reduced.

The invention discloses a key value management system and method based on salt stack and Redis. The system comprises a managed node, a database node and a storage node. The database node is a node which runs Redis primary and secondary clusters; salt stack client sides require to be deployed on servers of a Redis primary library and a Redis secondary library; the storage node stores Redis key value data, Redis key value recovery operation is facilitated, and the salt stack client sides are deployed; the managed node stores Redis primary and secondary cluster meta information, key value back-upmeta information, storage node meta information and database node meta information, and Web application service and a salt stack management end are deployed; and corresponding commands are transmitted to the salt stack client sides of the storage node through the salt stack management end, key value back-up files are transmitted, and thus, back-up and recovery operation is simplified. Examination, back-up, deleting and recovery operations of single or multiple key values are supported. The method is simple to operate, and one-touch operation can be implemented after an interface of a platformis filled with a small amount of information. Safe and reliable operation is implemented, the working efficiency is improved, and the manual maintenance cost is reduced.