31results about How to "Work in real time" patented technology

The invention discloses a single-frequency network radar multi-target tracking method. The method comprises a measurement interconnection module and a sequence tracking module. The measurement interconnection includes selecting a certain number of measurement and sending-receiving pairs to build low-dimensional association assumptions for association, quickly determining all of the association assumptions to remove incorrect association assumptions as for as possible, sorting association assumptions accepted in the last step, and solving to obtain second measurement after measurement interconnection on the basis of an overall association 0-1 integer programming model built according to sorted data in a single-frequency network. The sequence tracking is achieved based on the second measurement through Kalman filtering. The single-frequency network radar multi-target tracking method can be used for solving association fuzziness of measurement, a transmitting station and a target of a radar in the single-frequency network, greatly reduces the computation complexity by combining a heuristic method and an optimization model, achieves the multi-target tracking performance close to the known case of the association relationship and has popularization and application values.

The invention relates to a router temperature detection and heat radiation controlling circuit structure, which comprises a temperature sensor, a first comparator, and a heat radiation fan. The temperature sensor is arranged on the housing of the router. One end of the temperature sensor is connected with input voltage and the other end is connected with the first input end of the first comparator. T The first reference voltage is input into the second input end of the first comparator. The output end of the first comparator is connected with the heat radiation fan. The heat radiation fan is arranged at the bottom of the router. According to the structure of the invention, a heat radiation fan and a temperature sensor are arranged on the surface of a router. Through the comparison of the first reference voltage and the voltage at the other end of the temperature sensor, the opening and closing of the heat radiation fan can be controlled automatically. Without manual operations, the circuit structure can be conveniently used with the result that the heat radiation fan not only increases the heat radiation effect through the use of a heat radiation groove, but also ensures that the router works stably.

The invention relates to a multi-energy-source fuel cell vehicle thermal management system and a control method thereof. The fuel cell vehicle thermal management system comprises a fuel cell subsystem, a storage battery subsystem, a super capacitor subsystem, a heat exchange subsystem and a post-processing subsystem. The control method comprises the steps of detecting the output current and the output power of each energy source, the cooling liquid outlet temperature of each energy source and the heat required by a passenger compartment, and switching the control method of the heat managementsystem among an extremely cold mode, a small cycle, a medium cycle, a large cycle and an extremely hot mode according to the parameters. In each mode, the thermal management system controller realizesthe optimal efficiency of the whole power source part by controlling the flow of the cooling liquid passing through each subsystem, and the quick response of the whole power source part needs to be considered in a low-temperature environment, so that the economy of the fuel cell vehicle and the quick response performance in the low-temperature environment are improved.

The invention discloses a bill number identification method and system based on a deep neural network, and the method comprises the steps: sequentially carrying out the graying processing, binarization processing, normalization processing and two-dimensional image mapping one-dimensional data processing of an obtained two-dimensional mixed-shot bill original image, and greatly reducing the data size while maintaining the image information. Time waste caused by data transmission is reduced, and user experience is improved; by designing an OCTC model, introducing a one-dimensional convolution operation to perform feature extraction and image category calculation on image data, and using a plurality of small-size convolution kernels to increase a model receptive field, the experimental effect is ensured, the model parameter quantity is reduced, the effect of a lightweight model is realized, and storage and use by a user are facilitated; according to the method, the user can receive the judgment result of the number of bills in the image while uploading the image, the user experience of a mixed shooting bill recognition system is improved, meanwhile, the system is helped to carry out subsequent bill target detection and content recognition tasks, and real-time work of the system is achieved.

The invention relates to a vehicle real-time data acquisition system, a positioning verification method, and a mileage analysis method. The embodiment of the invention is advantageous in that: vehicle information captured by a front-end acquisition subsystem of a related time node is compared with the vehicle information uploaded by a vehicle-mounted acquisition subsystem at a cloud platform system, so as to judge whether the vehicle-mounted acquisition subsystem works abnormally or not, and ensure that the vehicle-mounted acquisition subsystem can effectively work in real time, accounting data of the cloud platform system is more accurate, and comprehensive control of vehicle operation data is promoted; by analyzing mileage of vehicles, working states of the vehicle-mounted acquisition subsystem are verified, and the travel charging of the vehicles is facilitated.

The invention discloses a multifunctional solid waste disposal cutter disk. The multifunctional solid waste disposal cutter disk comprises a rotating shaft and a second cutter disk. The lower end of the rotating shaft is provided with a rotating wheel. A first cutter disk is arranged below the second cutter disk. A third cutter disk is arranged above the second cutter disk. A fourth cutter disk isarranged above the third cutter disk. The two ends of the fourth cutter disk are both provided with blades. One side of the fourth cutter disk is provided with a separation disk. One side of the first cutter disk is provided with a silencer. The outer surface of an outer shell is provided with a water inlet and a water outlet. The inner surface of the outer shell is provided with a water storagetank. The multifunctional solid waste disposal cutter disk is provided with a rotating speed sensing mechanism, the separation disk and the water storage tank, the working rotating speed of a machinecan be detected in real time, the situation that solid waste enters the parts between every two adjacent cutter disks and the friction force is increased can also be prevented, the temperature in thecutting process is reduced, accidents are prevented, the multifunctional solid waste disposal cutter disk is suitable for different working conditions, and better application prospects are achieved.

The invention discloses a method for correcting a time error between sub converters of a double-channel analog-to-digital converter on the basis of machine learning, which comprises the steps of: carrying out chopping on a total output of the analog-to-digital converter, then obtaining a signal with a phase different of 90 degrees from a chopping signal by a Hilbert filter, then multiplying the signal by an original analog-to-digital converter output signal to obtain a feedback signal component closely related to the error, and extracting a corresponding error component to obtain an error signal; injecting the error signal into a feedback loop, according to taylor expansion correlation calculation, obtaining a corresponding deviation of sampling time among samplers; and adding the deviation of the signals, which is caused by a time error among the plurality of samplers, into an original acquired signal, carrying out correction on the error and completing recovery of data.

The invention discloses an asynchronous processing method and device based on a neural network accelerator, and the method comprises the steps: generating an asynchronous request through a main thread according to an algorithm service demand, generating a corresponding handle number, generating a group of command words according to the generated service request, adding a command word queue, and querying the command word queue, and if only one group of command words exist currently, setting the group of command words to be in an execution state, starting a neural network accelerator hardcore, otherwise, setting the group of command words to be in an initial state and returning. And after the execution of the hard core is finished, generating hardware interruption, starting soft interruption, querying a command word queue, taking out command words from the queue if the command words exist, judging whether the command words are in an initial state, if so, setting the command words to be in an execution state and starting the hard core, and otherwise, deleting the group of command words from the queue. And the main thread queries whether the hard core of the operation is completed or not according to the handle number. According to the operation, the asynchronous operation process of a plurality of network algorithms is realized, the asynchronous operation is processed through hard core interruption, the network performance is accelerated, and the bandwidth occupied by a processor is reduced.

The invention discloses a logistics sorting system. The system is characterized by comprising a recognition unit, an archival storage unit, a memory study unit, a control unit, a sorting device, a conveying device and a sorting way port; an intelligent algorithm is adopted, particularly, data a is input, a result y is output, the relation among a sorting algorithm F, a search algorithm S and knowledge analysis A is weighed, finally, output is determined through calculation recommendation, t is reaction time, and the formula: sigmaY=F(S(a)&A(a))%*t is met. The limit for packaged object specifications is greatly lowered, normalization of logistics information is greatly improved, the logistics distribution error rate is reduced, the defect that no corresponding distribution scheme exists inlogistics information transfer management is well solved, the self learning effect of real-time work and real-time updating is achieved, the accuracy of the logistics distribution scheme is greatly improved, and the sorting efficiency can be greatly improved by introducing the algorithm formula.

The invention discloses an open-circuit protection circuit of a solid-state power amplifier. The open-circuit protection circuit comprises a radio frequency link unit, a direct-current short-circuit antenna and a detection and control unit, the radio frequency link unit is electrically connected with the direct-current short-circuit antenna, and a connection line between the radio frequency link unit and the direct-current short-circuit antenna is externally connected with a detection and control unit; the detection and control unit comprises a high-frequency choke circuit, a pull-up circuit, a comparison circuit and an amplifier protection circuit; one end of the high-frequency choke circuit is respectively connected with the blocking capacitor and the direct-current short-circuit antenna, the other end of the high-frequency choke circuit is connected with the pull-up circuit, a connection point A in a connection line of the high-frequency choke circuit and the pull-up circuit is also connected with the comparison circuit, the comparison circuit is connected with the amplifier protection circuit, and the amplifier protection circuit is connected with the solid-state power amplifier. The problem of failure of the solid-state power amplifier caused by abnormal connection of an antenna, improper blind insertion and the like in array application of the solid-state power amplifier is solved.

The invention relates to a control system of a Delta converter-based UPS (Uninterrupted Power Supply) and belongs to the power supply control field. A new control method and topological structure areprovided for problems appearing in the work process of the system, and the instantaneous reactive power compensation effect and direct-current side energy balance control effect of a Delta converter are realized. The Delta conversion-based UPS (Uninterrupted Power Supply) has an uninterrupted power supply function and a power quality compensation function and can provide stable and efficient powerfor a load. The Delta conversion type UPS can act as a power quality compensator so as to compensate harmonic components, unbalanced components and reactive components contained in a power grid and the load. When the power outage of the power grid occurs, a main converter can achieve uninterrupted switching and full power operation of the power supply of a battery and the power supply of the power grid. The control system has the advantages of small power loss and high work efficiency.