33results about How to "Parallel implementation" patented technology

The present invention discloses a method for 4*4 sub-macroblock parallel intraframe prediction in H.264 / AVC, comprising the steps of: a unified intraframe prediction formula, a reference value array, a reference position table and a concrete implementation step of parallel intraframe prediction. The unified intraframe prediction formula is improved based on the feature of a CUDA and an intraframe prediction formula, by the corresponding prediction formulas of 9 prediction modes to be transformed into one formula to meet the requirements of the CUDA multi-thread single instruction multiple data stream, to achieve the fine-grained parallel in the intraframe prediction sub-macroblock. The reference value array and the reference position table are designed to cooperate with the unified intraframe prediction formula and to completely eliminate a large number of branch statements that affects the parallel algorithm performance. The present invention in the process of the intraframe prediction achieves the pixel level parallel, can effectively use multi-core resources in a GPU to accelerate the process of intraframe prediction, and shorts a coding time.

The invention discloses a shunt-wound rechargeable battery heavy constant current discharging device. The device is characterized by comprising one or more constant current discharging modules which are connected with one another in parallel by communication lines, wherein each discharging module is connected with a rechargeable battery; and the charging modules realize the coarse adjustment and fine adjustment of a discharging current by controlling the number of switches and regulating a given value of current fed to a DC converter. The device ensures high discharging current control and adjustment accuracy, has a simple structure, and can realize parallel connection among the discharging modules with different capacities and different current distribution modes such as an automatic current distribution mode, a manual current distribution mode and the like in the parallel connection.

The invention discloses a low clearance tunnel rigid overhead line system cantilever supporting device. The device comprises a supporting base (1) fixedly mounted on a tunnel wall, wherein the supporting base (1) is movably connected with a support frame (2); the support frame (2) is arranged slantly along a tunnel vault; the end of the support frame (2) is connected with an insulating part (3); the insulating part (3) is connected with an adjusting support (4); a suspension clamp (5) capable of moving on the adjusting support (4) for adjusting is arranged on the adjusting support (4). The support frame is slantly arranged along the tunnel vault, so that the whole set of device is arch-shaped, the mounting space of a low clearance tunnel can be met, the device also can reach or even be higher than the function of the conventional cantilever supporting device, the structure is simple, and the mounting technology is simplified.

The invention relates to a microgrid intelligent and rapid paralleling splitting device which comprises an IGBT (insulated gate bipolar transistor) switch with rapid on / off performance, a circuit breaker, a fuse, a voltage detection device, a current detection device, an IGBT temperature detection device, a circuit conditioning a sampled signal, a signal processing unit for processing and logical analysis and judgment, a driving circuit generating a main circuit switch device driving signal, and an LCD (liquid crystal display) screen used for displaying output information of the signal processing unit. The device realizes seamless switching between a microgrid and a large grid; the signal processing unit can be communicated with a microgrid controller; better coordinated operation between the microgrid and the large grid can be realized conveniently; and optimal management of a system is facilitated.

The embodiment of the invention provides a parallel implementation method, a device and a system for advanced motion vector prediction AMVP. The parallel implementation method includes that: a parallel area corresponding to parallel control parameters is obtained according to the parallel control parameters set by a coding terminal user; motion vector prediction selection sets of all the interframe units are constructed in a parallel way by the interframe prediction units of a coding unit in the parallel area corresponding to the parallel control parameters; motion vector prediction is carried out in a parallel way by all the interframe prediction units in the coding unit with utilizing the motion vector prediction selection sets constructed by the interframe prediction units, and predicted motion vectors are generated separately; motion estimation is carried out by each interframe prediction unit according to the predicted motion vector of the interframe prediction unit separately. In the parallel implementation method, the device and the system for advanced motion vector prediction AMVP, motion vector prediction selection sets of all the interframe units are constructed in a parallel way by the interframe prediction units in the parallel area corresponding to the parallel control parameters, thus parallel of motion estimation for prediction units of the coder end is achieved.

The invention provides a multi-cavity piezoelectric air-liquid pump. The air-liquid pump comprises a piezoelectric air pump body and a piezoelectric liquid pump body, and a piezoelectric air pump baseis in threaded connection with the piezoelectric liquid pump body; and the piezoelectric air pump body comprises a threaded pump cover I, a sealing ring I, three piezoelectric vibrators I and a pumpbody I, the three piezoelectric vibrators I and the pump body I form a pump cavity and divide the pump cavity into a first pump cavity body, a second pump cavity body and a third pump cavity body, follow-up thin wafer valves are arranged in corresponding inlets and outlets of the first pump cavity body and the second pump cavity body, a first fixing frame is arranged at the bottom of the pump bodyI, and a first V-shaped bluff body is arranged on the first fixing frame. According to the air-liquid pump, the follow-up thin wafer valves are quick in response, small in opening pressure and good in pump valve following performance, and in cooperation with the disc-shaped vibrators, the pumping performance of the parallel pump is greatly improved; and a valve pump, a valveless pump, the air pump body and the liquid pump body are integrated into a whole, all the pumps can be jointly controlled or independently controlled, simultaneous pumping and independent pumping of gas and liquid can berealized, the flow range is wide, and various working requirements can be met.

The invention relates to a parallel circuit for an AC power generator, and the parallel circuit comprises a main power generator, a slave power generator, two parallel control modules, and an inverter bridge, wherein the parallel control modules are used for controlling the power supply of the two power generators through the adjustment of a reference voltage and a proportionality coefficient of proportional links. Each parallel control module comprises a diode power module, a comparison unit, the proportional links, a PWM generator, an IGBT switching tube, a capacitor C1, and a diode. The output ends of the two power generators are respectively connected with the input ends of the corresponding parallel control modules. The output ends of the parallel control modules of the two power generators are connected to one common branch current bus. Finally, a steady alternating current is outputted through the inverter bridge. The parallel control modules of the two power generators are the same, and there is not line between the parallel control modules of the two power generators. The parallel control modules of the two power generators are connected to the same DC bus. The parallel circuit can solve the problems that a conventional power generator is large in parallel ring current, is complex in operation and is not balanced in power distribution.

The invention discloses a video-based fast template matching GPU implementation method which comprises the following steps: (1) copying image data to GPU equipment, wherein the operation of copying the images to the equipment and image template matching are executed at the same time as two independent streams, one of which executes a current-frame image template matching calculation process operation while the other of which executes a next-frame image copying operation; (2) storing the image template data in a shared memory, storing the image data in a texture memory, and randomly accessing by virtue of the texture memory for computation of match degree with the image template; (3) carrying out calculation at the same time by virtue of a multi-thread parallel calculation mode, wherein one thread calculates the template match degree quantity value of one position; and (4) determining the maximum value or the minimum value of the template match degree quantity value by virtue of a global atom and a shared atom so as to obtain a template matching result. According to the method disclosed by the invention, the template matching time can be obviously shortened and the practical value of a template matching algorithm can be improved.

The invention discloses a processing device and method for producing flat yarns. The device comprises an input roller used for inputting core yarns, a parallel arrangement yarn guide arranged above the input roller, a ring arranged above the parallel arrangement yarn guide and used for accommodating covering yarns, a flat yarn guide hole type yarn guide arranged above the ring, and an output roller arranged above the flat yarn guide hole type yarn guide, wherein the parallel arrangement yarn guide is provided with at least three first yarn guide holes arranged in parallel, the ring is provided with a ring center hole, the flat yarn guide hole type yarn guide is provided with a second yarn guide hole, the first yarn guide holes are matched with the ring center hole, and the ring center hole is matched with the second yarn guide hole. By the adoption of the structure, a plurality of core yarns can pass through yarn guiding devices arranged in parallel so as to enter a twisting system in parallel; meanwhile, the flat yarn guide hole type yarn guide is adopted for an output yarn guiding device, so that it is guaranteed that the core yarns are arranged in parallel when fixed by the covering yarns.

The invention discloses automatic road marking pre-marking equipment and a working method. The automatic road marking pre-marking equipment comprises a self-walking trolley, a laser radar, an external interaction terminal, a GNSS fixed station and a GNSS mobile station, wherein the laser radar and the GNSS mobile station are fixed on the self-walking trolley; the external interaction terminal is in interactive communication with the self-walking trolley, the laser radar, the GNSS fixed station and the GNSS mobile station; the GNSS fixed station and the GNSS mobile station carry out interactive communication, and accuracy of pre-scribing positioning is guaranteed. According to the positioning information and the set scribing rule, the arranged spraying head can conduct automatic spraying and scribing according to the scribing path according to the scribing rule, automatic pre-scribing is achieved, labor cost is reduced, and pre-scribing efficiency is improved.

The invention discloses a filter fixing member, a manufacturing method of the filter fixing member and a manufacturing method of a compact-type coarse wavelength division multiplexer. The filter fixing member comprises a top surface, a bottom surface opposite to the top surface, and two side surfaces which are connected with the top surface and the bottom surface and are arranged in a mode of being parallel and opposite to each other. The top surface is equipped with multiple optical path grooves which extend along a preset optical path of the compact-type coarse wavelength division multiplexer to form multiple optical path passages. The multiple optical path grooves penetrate the two side surfaces, so that multiple optical path interfaces are formed in the two side surfaces. The two side surfaces are mounting surfaces used for mounting optical devices. Through the filter fixing member disclosed by the invention, a problem that requirements for the side sag of an optical device when side surfaces of the optical device and a substrate are bonded are too high can be avoided so as to improve the qualified rate, a complex operation of precisely adjusting angles of the optical device is removed in production, the operation difficulty is reduced, and the working time is saved.

The invention relates to an automatic detection, automatic dispensing and automatic sorting device for LED lamp beads arranged in an array, which comprises a chip detection device for detecting whether each chip in the LED lamp beads arranged in the array works normally or not and a dispensing device for dispensing the chips in the LED lamp beads arranged in the array, the dispensing device comprises a plurality of dispensing machines arranged in an array, and the chip detection device is used for detecting whether each chip in the LED lamp beads arranged in the array works normally or not. Each dispensing machine corresponds to one LED lamp bead needing to be dispensed, the chip detection equipment transmits a working state detection result of each chip in the LED lamp beads which are arranged in an array to the control system, and the control system recognizes the chips which work abnormally and coordinate information corresponding to the chips; the system controls the electric valve of the dispensing machine corresponding to the coordinates of the chip to be closed so as to stop dispensing on the abnormally working chip, the system further comprises a sorting device electrically connected with the control system, and the control system subsequently controls the sorting device to remove the abnormally working chip. The device can save fluorescent powder, glue and manpower.

The invention provides a power device packaging structure and a preparation method thereof. The power device packaging structure comprises a plurality of power device packaging structure units, the power device packaging structure unit comprises a first heat dissipation substrate, a power device, a diode and a high-heat-conductivity plastic packaging material shell, wherein the power device and the diode are arranged on the surface of the first heat dissipation substrate, and the upper part of the power device packaging structure unit is subjected to plastic packaging by using a high-heat-conductivity plastic packaging material, a high-heat-conductivity plastic packaging material shell. The first heat dissipation substrate is in a long plate shape, and the power device and the diode are sintered on the surface of the first heat dissipation substrate through a sintering technology. The power device packaging structure based on the sintering technology can realize series-parallel connection in a two-dimensional plane, has good heat conductivity and reliability, and is simple in grid control; and in the manufacturing of a single subunit, the chip surface is not required to be directlycontacted, so that the risk of chip damage caused by uneven pressure is avoided.

The invention relates to over-snow transporting equipment, in particular to a sleigh suspension mechanism. The sleigh suspension mechanism comprises a supporting seat, connecting rods, sleigh boards and shock absorbers. Each connecting rod is hinged between the supporting seat and the corresponding sleigh board. A parallelogram mechanism is formed among the supporting seat, the sleigh boards and the connecting rods. Each shock absorber is further arranged between the supporting seat and the corresponding sleigh board and comprises a shock absorber sliding shaft, a spring and a shock absorber guide sleeve. One end of each shock absorber sliding shaft and one end of the corresponding shock absorber guide sleeve are hinged to the corresponding sleigh board and the supporting seat. The other end of each shock absorber sliding shaft and the other end of the corresponding shock absorber guide sleeve can be connected in a relatively telescopically inserted manner. Each shock absorber sliding shaft and the corresponding shock absorber guide sleeve are sleeved with the corresponding spring. The two ends of each spring are connected with the corresponding shock absorber sliding shaft and the corresponding shock absorber guide sleeve. By the adoption of the sleigh suspension mechanism, the sleigh boards and the ground are kept in parallel, the pressure distribution evenness of the bottom face of each sleigh board is guaranteed, the adaptive capacity of a sleigh to the ground is improved, and the sleigh suspension mechanism is suitable for a vehicle body mechanism using the sleigh as a moving manner in an ice and snow environment.

The invention discloses an integrated low-voltage energy storage power supply system for a vehicle. The integrated low-voltage energy storage power supply system comprises a storage battery; a power supply main switch, the input end of the power supply main switch being connected with the positive electrode of the storage battery, and the negative electrode of the storage battery being the whole vehicle common ground; a charging relay, the input end of the charging relay being connected with the output end of the power supply main switch, and the output end being grounded; a current-limiting resistor, the input end of the current-limiting resistor being connected with the output end of the charging relay; a super capacitor module, formed by connecting a plurality of super capacitors in series and in parallel; a parallel relay; and a starter, the output end of the starter being connected with a whole vehicle engine starter. The parallel relay enables the output end of the power supply main switch to be connected in parallel with the positive electrode of the super capacitor module. The input end of the starter is connected with the output end of the current-limiting resistor, the output end of the parallel relay and the positive electrode end of the super capacitor module. Charging current-limiting control can be realized, and the starting success rate is improved. The invention further provides a control method of the integrated low-voltage energy storage power supply system for the vehicle.

The invention discloses a clamping device. The device comprises a base plate and a plurality of movement execution mechanisms. Each movement execution mechanism comprises a first direction movement device, a second direction movement device and a third direction movement device, wherein the first direction movement device is arranged on the base plate, the second direction movement device is arranged on a work portion of the first direction movement device, and the third direction movement device is arranged on a work portion of the second direction movement device; and a clamping mechanism isarranged on a work portion of the third direction movement mechanism and comprises a clamping portion, and the clamping portion comprises a plurality of clamping jaws and a drive device used for driving the clamping jaws. The multiple movement execution mechanisms perform multithreading parallel operation.

The invention discloses a JPEG decoding method and device, which are applied to the field of image processing, and are characterized in that an FPGA accelerator card reads M pictures from a DDR of theFPGA accelerator card, and M is an integer greater than 1; the FPGA acceleration card correspondingly allocates M pictures read from the DDR to M JPEG decoders, and the M JPEG decoders share the sameGMEM resource on the FPGA acceleration card to perform M paths of parallel JPEG decoding on the M pictures to obtain M paths of decoded data streams; the FPGA acceleration card reads and converges the M paths of decoded data streams to obtain converged decoded data; and the converged decoded data is output to the DDR. According to the invention, the JPEG decoding efficiency is improved.

The invention discloses a parallel structure with multiple sets of embedded touch-control liquid crystal displays, which includes a display installation limiting mechanism and a bearing outer frame fixing mechanism, and both sides of the display installation limiting mechanism are movably connected to the bearing outer frame fixing mechanism. When installing multiple groups of displays in this parallel structure, they can be respectively put into the corresponding display installation limit card slots, and the first wire connection hole and the second wire connection hole can be used to connect wires to each other to form a parallel connection. The structure realizes the parallel connection of AC constant voltage sources, and because of the parallel shunt, when the inflowing current is too large, it can be shunted, and it is not easy to burn out. Even if a display branch is damaged and disconnected, the other display will still form a path with the main circuit and maintain normal operation. Compared with the general series connection method, the safety is also higher.

The invention discloses a processing device and method for producing flat yarns. The device comprises an input roller used for inputting core yarns, a parallel arrangement yarn guide arranged above the input roller, a ring arranged above the parallel arrangement yarn guide and used for accommodating covering yarns, a flat yarn guide hole type yarn guide arranged above the ring, and an output roller arranged above the flat yarn guide hole type yarn guide, wherein the parallel arrangement yarn guide is provided with at least three first yarn guide holes arranged in parallel, the ring is provided with a ring center hole, the flat yarn guide hole type yarn guide is provided with a second yarn guide hole, the first yarn guide holes are matched with the ring center hole, and the ring center hole is matched with the second yarn guide hole. By the adoption of the structure, a plurality of core yarns can pass through yarn guiding devices arranged in parallel so as to enter a twisting system in parallel; meanwhile, the flat yarn guide hole type yarn guide is adopted for an output yarn guiding device, so that it is guaranteed that the core yarns are arranged in parallel when fixed by the covering yarns.

The invention discloses a clamping device. The device comprises a base plate and a plurality of movement execution mechanisms. Each movement execution mechanism comprises a first direction movement device, a second direction movement device and a third direction movement device, wherein the first direction movement device is arranged on the base plate, the second direction movement device is arranged on a work portion of the first direction movement device, and the third direction movement device is arranged on a work portion of the second direction movement device; and a clamping mechanism isarranged on a work portion of the third direction movement mechanism and comprises a clamping portion, and the clamping portion comprises a plurality of clamping jaws and a drive device used for driving the clamping jaws. The multiple movement execution mechanisms perform multithreading parallel operation.

The invention discloses a video-based fast template matching GPU implementation method which comprises the following steps: (1) copying image data to GPU equipment, wherein the operation of copying the images to the equipment and image template matching are executed at the same time as two independent streams, one of which executes a current-frame image template matching calculation process operation while the other of which executes a next-frame image copying operation; (2) storing the image template data in a shared memory, storing the image data in a texture memory, and randomly accessing by virtue of the texture memory for computation of match degree with the image template; (3) carrying out calculation at the same time by virtue of a multi-thread parallel calculation mode, wherein one thread calculates the template match degree quantity value of one position; and (4) determining the maximum value or the minimum value of the template match degree quantity value by virtue of a global atom and a shared atom so as to obtain a template matching result. According to the method disclosed by the invention, the template matching time can be obviously shortened and the practical value of a template matching algorithm can be improved.

The invention discloses a circuit breaker assembly and a single P circuit breaker. The single P circuit breaker comprises a shell and an input end installed on the shell. The shell is provided with a through hole, and a copper bar can be inserted into the through hole. The input end comprises a current conduction piece positioned in the shell; the copper bar which is installed on the shell, wherein the copper bar and the current conduction piece are overlapped, and one end of the copper bar can extend out of the shell; a pressing mechanism which is installed on the shell and tightly presses an overlapped part in the copper bar which is overlapped with the current conduction piece in the shell on the current conduction piece in a pressing state. When the parallel connection of input ends of a plurality of single P circuit breakers is needed, adding a cable for parallel connection, a parallel copper bar and the like are not needed, the parallel connection of a plurality of single P circuit breakers can be realized through the copper bar, the current conduction piece and pressing mechanism of each single P circuit breaker, and a circuit space in the parallel connection of a plurality of single P circuit breakers can be saved.

The invention provides a loading method for a high-speed grinding machine for valve rod machining. The loading method for the high-speed grinding machine for valve rod machining is capable of solvingthe technical problem that a loading speed is difficultly increased when a valve rod is machined through a numerically-controlled lathe. The loading method for the high-speed grinding machine for valve rod machining is characterized by comprising the following steps: 1, clamping the valve rod; 2, transferring the valve rod to the front end of a clamp of a workpiece shaft and coaxially arranging the valve rod with the workpiece shaft; 3, in the rotating state of the workpiece shaft, jacking the valve rod into the clamp at the front end of the workpiece shaft; and 4, clamping the valve rod by the clamp. With the adoption of the method, on one hand, the workpiece shaft drives the valve rod to synchronously rotate, the valve rod in the rotating state is capable of ensuring parallelism with theaxis of the workpiece shaft, that is, capable of realizing relative location for the valve rod in the clamp, so that the clamping accuracy can be guaranteed; and on the other hand, the start and stopof the workpiece shaft can be avoided, and a loading speed for the valve rod is effectively increased, so that the machining efficiency for the valve rod is increased.

The invention discloses a parallel generation method of a random two-phase medium model based on a GPU and electric equipment. The method comprises the steps that a CPU reads initial distribution information corresponding to two-phase media; the CPU defines a target two-point probability function and a target linear path function; the GPU generates an initial random model; the CPU generates a current two-point probability function and a current linear path function, and a combined error is calculated; the CPU judges whether the combined error is smaller than a target value or not; if the combined value is smaller than the target value, the CPU draws the current initial random model, and the current initial random model is stored as the final two-phase medium model; if the combined error isnot smaller than the target value, the iteration step is executed repeatedly until the combined error is smaller than the target value, then the CPU draws the current initial random model, and the current initial random model is stored as the final two-phase medium model. The parallel generation method is high in computing speed, low in computing cost, low in hardware cost and convenient and easyto use.

The invention provides a parallelism adjusting method which comprises the steps that 1, a base is installed on a reference plane, 2, a second adjusting block is installed on the base, and the lower portion of the second adjusting block is movably connected with the base through bolts, 3, adjusting screws are installed at the two ends of the second adjusting block, 4, the upper end of the second adjusting block is movably connected with a first adjusting block through bolts, the rotating direction of the first adjusting block is enabled to be perpendicular to that of the second adjusting block; 5, adjusting screws are installed at the two ends of the first adjusting block; 6, an object to be measured is fixed to the first adjusting block; 7, the height of two ends of the first adjusting block is measured, and the height of two ends of the first adjusting block is adjusted to be same; 8, the height of two ends of the second adjusting block is measured, and the height of two ends of the second adjusting block is adjusted to be identical. According to the method, the parallelism of the two directions is adjusted by adjusting the two adjusting blocks, finally, the plane of the object to be measured is parallel to the reference plane, the structure is simple, and adjustment is convenient.

The invention relates to a parallel circuit for an AC power generator, and the parallel circuit comprises a main power generator, a slave power generator, two parallel control modules, and an inverter bridge, wherein the parallel control modules are used for controlling the power supply of the two power generators through the adjustment of a reference voltage and a proportionality coefficient of proportional links. Each parallel control module comprises a diode power module, a comparison unit, the proportional links, a PWM generator, an IGBT switching tube, a capacitor C1, and a diode. The output ends of the two power generators are respectively connected with the input ends of the corresponding parallel control modules. The output ends of the parallel control modules of the two power generators are connected to one common branch current bus. Finally, a steady alternating current is outputted through the inverter bridge. The parallel control modules of the two power generators are the same, and there is not line between the parallel control modules of the two power generators. The parallel control modules of the two power generators are connected to the same DC bus. The parallel circuit can solve the problems that a conventional power generator is large in parallel ring current, is complex in operation and is not balanced in power distribution.

The invention provides a serial-parallel tuned inertial damper, which includes: a shell body, a shearing rotating member and a screw rod, the shell body has a cavity, and an inner support sleeve is provided at the bottom of the cavity, and the shell An accommodating space is formed between the inner wall of the body and the inner support sleeve to contain the viscous medium; the shearing rotating member is rotatably arranged in the cavity and above the inner supporting sleeve, and the shearing rotating member is provided with a rotating slice to Shearing the viscous medium produces a damping effect, the screw rod is passed through the shell body and extends into the sleeve, and is screwed with the shearing rotating member to convert the axial movement of the screw rod into the rotating motion of the shearing rotating member, Due to its own viscous characteristics, the viscous medium produces a drag force distributed over the entire circumference of the rotating slice, thus forming a large torque on the rotating shaft, and this torque is transformed into a large obstacle after shearing the rotating part. The damping force of the axial movement of the screw rod acts on the engineering structure.

The present invention discloses a method for 4*4 sub-macroblock parallel intraframe prediction in H.264 / AVC, comprising the steps of: a unified intraframe prediction formula, a reference value array, a reference position table and a concrete implementation step of parallel intraframe prediction. The unified intraframe prediction formula is improved based on the feature of a CUDA and an intraframe prediction formula, by the corresponding prediction formulas of 9 prediction modes to be transformed into one formula to meet the requirements of the CUDA multi-thread single instruction multiple data stream, to achieve the fine-grained parallel in the intraframe prediction sub-macroblock. The reference value array and the reference position table are designed to cooperate with the unified intraframe prediction formula and to completely eliminate a large number of branch statements that affects the parallel algorithm performance. The present invention in the process of the intraframe prediction achieves the pixel level parallel, can effectively use multi-core resources in a GPU to accelerate the process of intraframe prediction, and shorts a coding time.