Abstract: An air conditioning system includes a main circuit having a multi-stage compressor, a condenser, a throttling element and an evaporator connected by pipelines; and a cooling branch, the inlet of which is connected to the main circuit between the condenser and the throttling element, and the outlet of which is connected to at least one of the first-stage suction port and the intermediate-stage suction port of the multi-stage compressor, wherein the cooling branch flows through the drive motor of the multi-stage compressor, and a regulating valve for controlling the opening of the cooling branch is provided on the cooling branch; and a control module that controls the opening of the regulating valve on the cooling branch based on the temperature of the outlet downstream of the drive motor on the cooling branch and the intermediate suction pressure of the intermediate-stage suction port of the multi-stage compressor.

Abstract: A two-stage energy dissipation type shed tunnel support structure connected by a principle of Dougong and a design method thereof are provided. The two-stage energy dissipation type shed tunnel support structure includes a Dougong joint domain, crossbeams and columns. The Dougong joint domain includes section steel members, wavy-wall cylindrical elastoplastic buffers, U-shaped sliding connecting troughs and high-strength bolts. Multiple layers of the section steel members are orthogonally stacked to form a Dougong shape, the wavy-wall cylindrical elastoplastic buffers are arranged between adjacent layers of the section steel members, and the U-shaped sliding connecting troughs are arranged at upper and lower ends of the wavy-wall cylindrical elastoplastic buffers, which realizes a connection between the wavy-wall cylindrical elastoplastic buffers and a section steel in two orthogonal directions.

Abstract: The present disclosure provides a six-cylinder opposed free piston internal combustion engine generator. The generator comprises two free piston internal combustion engine sets, one opposed piston internal combustion engine set and two linear generator sets. Air entering cylinders is subjected to first-stage compression in low-pressure cylinder sets in the free piston internal combustion engine sets and the opposed piston internal combustion engine set and then subjected to second-stage compression in high-pressure cylinder sets, and a high pressure gas produced after the combustion is subjected to first-stage expansion in the high-pressure cylinder sets and then subjected to second-stage expansion in the low-pressure cylinder sets.

Abstract: Provided is a free piston generator based on a split thermodynamic cycle, which belongs to the technical field of power energy. The present disclosure solves the problem of low power generation efficiency of an existing free piston generator. The free piston generator includes a linear generator set and two internal combustion engine sets arranged at two ends of the linear generator set. Air is first subjected to first-stage compression by the low-pressure cylinder set in the internal combustion engine sets and is then subjected to second-stage compression in the high-pressure cylinders, so that the intake pressure of an internal combustion engine is effectively increased, which is favorable for increasing the average effective pressure in a work process, thereby improving the thermal efficiency and the power generation efficiency of the free piston generator.

Abstract: A free piston generator based on a rigid synchronous transmission system is provided, which belongs to the technical field of power energy. The present disclosure solves the problems of low power generation efficiency and low stability of the existing free piston generator. The free piston generator based on the rigid synchronous transmission system includes a first linear generator set, a second linear generator set, a rigid synchronous transmission assembly, two high-pressure cylinders arranged at two ends of the first linear generator set, and two low-pressure cylinders arranged at two ends of the second linear generator set.

Abstract: An example apparatus for mining multi-scale hard examples includes a convolutional neural network to receive a mini-batch of sample candidates and generate basic feature maps. The apparatus also includes a feature extractor and combiner to generate concatenated feature maps based on the basic feature maps and extract the concatenated feature maps for each of a plurality of received candidate boxes. The apparatus further includes a sample scorer and miner to score the candidate samples with multi-task loss scores and select candidate samples with multi-task loss scores exceeding a threshold score.

Abstract: The present application discloses a display panel, including a driving module including a clock control module and a common voltage signal line electrically connected to the clock control module, wherein the common voltage signal line is input with a differential square wave signal; and a display module including a color filter substrate and a common electrode layer disposed on the color filter substrate, and the common electrode layer is electrically connected to the common voltage signal line.

Abstract: A computer-implemented method includes obtaining, by one or more processing devices, a wireless message of an Internet of Things (IoT) device by monitoring a wireless signal sent by the IoT device; determining, by the one or more processing devices, a target running feature of the IoT device based on the wireless message; comparing, by the one or more processing devices, the target running feature with a historical normal running feature of the IoT device; and determining, by the one or more processing devices, whether the IoT device runs abnormally based on the comparing of the target running feature with the historical normal running feature of the IoT device.

Abstract: A centrifugal compressor and a refrigerating device. The compressor includes: a shell, which has a fluid inlet at a first position of the shell, and a fluid outlet at a second position of the shell, a motor assembly, which is arranged in the shell and includes a stator and a rotor, the rotor including a vertically arranged rotor shaft, and the rotor shaft including a lower end and an upper end; a centrifugal compression mechanism, an impeller of which is connected with the rotor shaft so as to be driven by the motor assembly; and a guide member, which is located above the centrifugal compression mechanism, and which defines a flow passage alone or together with a top part of the shell.

Abstract: The invention relates to a centrifugal compressor and a method of operating the same. A centrifugal compressor of the invention includes an impeller, a diffuser and a variable diffusion system, wherein the diffuser is disposed downstream of the impeller and has a respective diffuser frame that defines at least a channel through which the fluid flows from the impeller; moreover, the variable diffusion system comprises a movable body that is able to move relative to the channel and an electromagnetic actuating device, and the electromagnetic actuating device controls the movable body to move relative to the channel in an electromagnetic driving manner. The variable diffusion system of the centrifugal compressor in the invention is simple in implementation structure, low in cost, easy to operate and good in stability and reliability.

Abstract: An example apparatus for mining multi-scale hard examples includes a convolutional neural network to receive a mini-batch of sample candidates and generate basic feature maps. The apparatus also includes a feature extractor and combiner to generate concatenated feature maps based on the basic feature maps and extract the concatenated feature maps for each of a plurality of received candidate boxes. The apparatus further includes a sample scorer and miner to score the candidate samples with multi-task loss scores and select candidate samples with multi-task loss scores exceeding a threshold score.

Abstract: An example apparatus for mining multi-scale hard examples includes a convolutional neural network to receive a mini-batch of sample candidates and generate basic feature maps. The apparatus also includes a feature extractor and combiner to generate concatenated feature maps based on the basic feature maps and extract the concatenated feature maps for each of a plurality of received candidate boxes. The apparatus further includes a sample scorer and miner to score the candidate samples with multi-task loss scores and select candidate samples with multi-task loss scores exceeding a threshold score.

Abstract: The invention is a method for controlling execution of an application. The method comprising: installing and activating a software license unit including License terms and a secure repository comprising both an applet and parameters, providing a virtual USB dongle including a command gate, a License validator, a VM controller and a VM engine initially devoid of applet, verifying the License terms and only if the verification of the License terms is successful: loading said applet and parameters to the VM engine and enabling the Command gate, initializing configuration data and secret data in the VM engine by using the parameters stored in the VM engine then exchanging, between the applet and said hardware function driver, USB messages to control execution of said application.

Abstract: A computer-implemented method includes obtaining, by one or more processing devices, a wireless message of an Internet of Things (IoT) device by monitoring a wireless signal sent by the IoT device; determining, by the one or more processing devices, a target running feature of the IoT device based on the wireless message; comparing, by the one or more processing devices, the target running feature with a historical normal running feature of the IoT device; and determining, by the one or more processing devices, whether the IoT device runs abnormally based on the comparing of the target running feature with the historical normal running feature of the IoT device.

Abstract: Methods, apparatus, systems and articles of manufacture for loss-error-aware quantization of a low-bit neural network are disclosed. An example apparatus includes a network weight partitioner to partition unquantized network weights of a first network model into a first group to be quantized and a second group to be retrained. The example apparatus includes a loss calculator to process network weights to calculate a first loss. The example apparatus includes a weight quantizer to quantize the first group of network weights to generate low-bit second network weights. In the example apparatus, the loss calculator is to determine a difference between the first loss and a second loss. The example apparatus includes a weight updater to update the second group of network weights based on the difference. The example apparatus includes a network model deployer to deploy a low-bit network model including the low-bit second network weights.

Abstract: Methods, systems, apparatus, and articles of manufacture are disclosed to train a neural network. An example apparatus includes an architecture evaluator to determine an architecture type of a neural network, a knowledge branch implementor to select a quantity of knowledge branches based on the architecture type, and a knowledge branch inserter to improve a training metric by appending the quantity of knowledge branches to respective layers of the neural network.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for face verification are provided.

Abstract: The present disclosure discloses a combined offshore wind power foundation with duct piles and a bucket, which comprises duct piles and a bucket foundation; the bucket foundation includes a steel bucket skirt, a bucket top cover, main beams, ring beam and a transition section. A top of the steel bucket skirt and the bucket top cover are integrally formed into a semi-closed structure. The bucket top cover is provided thereon with main beams uniformly arranged in a circumferential direction. The ring beam are arranged at an edge of the bucket top cover. Each duct pile includes a duct, a steel pile, duct supports and high-strength mortar, each main beam is provided at its outer side end with one of corresponding. The duct support is connected with the main beam and the ring beam, and integrated with the bucket top cover, the main beams and the ring beam.

Abstract: Methods, systems, and devices, including computer programs encoded on computer storage media, for security scanning a mini program are provided. One of the methods includes: obtaining a target mini program to be released, invoking a security scanning strategy combination to perform multi-dimensional security scanning on the target mini program; and when the target mini program passes the multi-dimensional security scanning, releasing the target mini program to a server. The multi-dimensional security scanning may include malicious code scanning on the target mini program, security loophole scanning on the target mini program, and security loophole scanning on a server interface of the target mini program.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for face verification are provided.