49results about How to "Realize full coverage detection" patented technology

The invention discloses a speech quality detection method and device, which relates to the technical field of speech detection, and mainly aims at improving the comprehensiveness and accuracy of speech quality detection. The main technical scheme of the invention is characterized by converting the speech to be quality-checked into the text to be quality-checked; extracting quality inspection itemsone by one from a quality inspection item collection, wherein the quality inspection items are used for detecting whether the text to be inspected conforms to requirements; using a pre-trained machine learning model to detect whether the text to be inspected contains content matching with the quality inspection item; according to the matching result and the detection rule corresponding to the quality inspection entry, determining whether the speech to be inspected is qualified or not. The invention is used for the normative detection of the voice data content.

The invention discloses a plane material surface defect detection device which comprises a main body frame, a linear light source mounted on the main body frame and at least one detection module mounted on the main body frame, wherein the main body frame comprises two upright posts positioned on the two sides, and a crossbeam fixedly mounted on the two upright posts; a sliding guide rod is arranged on the crossbeam; the detection module comprises a mounting seat mounted on the sliding guide rod in a slide fit manner; a CCD image sensor and a point group type laser respectively positioned on the two sides of the crossbeam are mounted on the mounting seat; a linear light source mounting seat is respectively arranged on each of the two upright posts; two ends of the linear light source are respectively mounted on the two linear light source mounting seats; an adjusting mechanism used for adjusting the position of the linear light source in the direction perpendicular to the crossbeam is arranged between the linear light source mounting seats and the linear light source. The plane material surface defect detection device can realize defect detection of the surface of the plane material in a fully covering manner.

The invention discloses a phased array sound wave advanced prediction system and method suitable for drilling-blasting tunnels. The system includes sound wave transduction devices and a host system, and the sound wave transduction devices transmit and receive sound waves; a data temporary storage module stores sound wave reflection signal data and focused sound wave reflection signal data obtainedby three-dimensional detection; three-dimensional coordinates of the sound wave transduction devices are imported into an acquisition module, and the acquisition module calculates and controls the emission time and the excitation sequence of each sound wave transduction device according to the three-dimensional coordinates, so that delayed transmitting is achieved to form focused sound waves; a control module is used for temporarily storing the acquired data into a data temporary storage module and transmitting the data to a data processing imaging module; and the data processing imaging module applies a deep learning algorithm for automatic identification and judging of the boundary of the geological anomalous body, conducts processing, analysis and integration on the information to forma three-dimensional detection imaging result, and finally a high-resolution three-dimensional detection image in front of a tunnel face can be obtained.

The invention relates to a flaw detection support and an ultrasonic flaw detector, wherein the flaw detection support comprises a hanging frame assembly, a sliding frame assembly and a sliding shoe probe assembly; the hanging frame assembly comprises at least one hanging frame guide assembly, the hanging frame guide assembly comprises a guide shaft and a first spring, and the guide shaft is vertically arranged; the sliding frame assembly comprises a sliding frame plate assembly and at least one sliding frame guide assembly, the first spring and the sliding frame plate assembly are arranged onthe guide shaft in a sleeving mode, the sliding frame guide assembly comprises a guide shaft and a second spring, the guide shaft is horizontally fixed to the sliding frame plate assembly, and the second spring and the sliding shoe probe assembly are arranged on the guide shaft in a sleeving mode. The carriage assembly ensures that the skid shoe probe assembly is downwardly attached to a steel rail; the sliding shoe probe assembly can horizontally move along the guide shaft, so that the sliding shoe probe assembly is tightly attached to the steel rail from the side direction. The skid shoe probe assembly is always reliably attached to the preset flaw detection position of the steel rail along with the curve change of the inner side face of the steel rail.

The invention discloses an infrared probing system and method applied to a high speed train, and belongs to the field of railway safety monitoring. As for a probing system, a host is connected with acontrol box and a train number box, the control box is connected with a bottom probing box and two multielement probes, and each multielement probe comprises a multielement photonic device, a wide-angle multielement optical system and a probe processing circuit. The probing method comprises the steps that when a train passes through, infrared radiation is generated by a part at the train bottom, and is received by the wide-angle multielement optical systems and gathered to chips of the photonic devices, a voltage signal is generated through photovoltaic conversion, the probe processing circuits conduct treatment, such as amplification, filtering and regulating, on the voltage signal, after the voltage signal is subjected to treatment by the control box, the voltage signal is transmitted tothe host to generate a heat map, then correcting is conducted through an intelligent heat map processing module and a center large data platform of the system, and the integrated temperature Ts of each part is obtained; and whether the temperature Ts is larger than the preset value or not is judged, strong-heat early warning or heat shock alarming is sent according to the different preset values,and effective and accurate probing on the part at the train bottom of the high speed train is achieved.

The invention discloses a steel tube pulse magnetization magnetic flux leakage detecting device which comprises two pairs of circumferential magnetizers and an area array sensor. The two pairs of the circumferential magnetizers are vertically arranged in the circumferential direction of a steel tube, pulse magnetization currents of which the values of amplitude are equal and the difference of phase positions is a half cycle are introduced, and time-sharing magnetization of different parts of the steel tube is achieved; the area array sensor is arranged in the circumferential direction of the steel tube in a 360-degree full-coverage mode, and defect magnetic flux leakage fields of effective magnetization parts are obtained according to timing sequences of pulse currents. According to the steel tube pulse magnetization magnetic flux leakage detecting device, a movement mode that a steel tube to be detected is pushed in a high-speed linear mode is achieved, and full coverage detection of longitudinal defects of the steel tube can be achieved through a set of fixed type detection units. Meanwhile, the problem of longitudinal defect detection for long and thin ferromagnetic components such as coiled tubings, square drill rods and steel rails which can not do rotary motion by themselves easily is solved.

The embodiment of the invention discloses a long-distance optical fiber detection method, device, equipment and system and a storage medium. The method comprises the following steps: in response to adetection request of a target node on a to-be-detected optical fiber, determining a first sampling sequence and a second sampling sequence which are respectively formed by propagation of a first optical signal and a second optical signal which are respectively emitted from two ends of the to-be-detected optical fiber through an OTDR on the to-be-detected optical fiber; determining the total lengthof the to-be-detected optical fiber; and generating a detection result according to the first sampling sequence, the second sampling sequence and the total length, and sending the detection result tothe target node. The method comprises the following steps: determining a first sampling sequence and a second sampling sequence formed by respectively propagating a first optical signal and a secondoptical signal which are respectively emitted from two ends of the to-be-detected optical fiber through the OTDR on the to-be-detected optical fiber; and generating the detection result of the to-be-detected optical fiber by combining the total length of the to-be-detected optical fiber, so that full-coverage detection of the to-be-detected optical fiber can be realized when the optical fiber is too long, and the reliability of optical fiber detection is improved.

The invention discloses a hardware Trojan horse detection method based on deep learning. The method mainly solves the problems that an existing method is high in cost, high in omission ratio and low in efficiency. According to the scheme, the method comprises the steps of collecting image sets, and constructing two training sets; collecting a to-be-detected image set; training the residual channelattention network by using the first training set, training the cyclic consistency generative adversarial network by using the second training set, and sequentially sending the microscopic images inthe to-be-detected image set to the trained residual channel attention network and the cyclic consistency generative adversarial network to obtain images which are homologous with the mother set microscopic images; enhancing the images homologous with the mother set microcosmic images and the corresponding mother set microcosmic images, and carrying out binarization segmentation and denoising on the enhanced images; connected region marking being carried out on the denoised image, exclusive-OR operation being carried out on the denoised image, and the region with the operation result being 1 being a hardware Trojan horse. The method has the advantages of higher detection precision, higher speed and simpler operation, and can be used for preparing an integrated circuit chip.

The invention discloses a steel pipe longitudinal defect magnetic leakage detection device based on a rotary magnetizing field. The device comprises a rotary magnetizer, an excitation power supply, an area array sensor array, an area array sensor array support, a filter, an amplifier and an A/D (analog-to-digital) converter. The excitation power supply provides three-phase asynchronous alternating current for the rotary magnetizer, and the rotary magnetizing field is generated inside the rotary magnetizer. The annular area array sensor array is fixed into the rotary magnetizer through the support and fully circumferentially covers a steel pipe. When the steel pipe passes the detection device, the rotary magnetizing field fully, circumferentially and uniformly magnetizes the steel pipe. If the steel pipe has a longitudinal defect, a defect magnetic flux leakage field is generated, picked by a sensor at a position corresponding to the defect and converted into an electric signal, the electric signal is transmitted to a computer sequentially through the filter, the amplifier and the A/D converter, and the longitudinal defect is detected. The steel pipe is fully, circumferentially and uniformly magnetized by the device, longitudinal defect detection efficiency is improved, and cost is reduced.

The invention relates to the technical field of automatic nondestructive testing, in particular to an electromagnetic sound detection robot for a storage tank. The robot comprises an electromagnetic sound sensor, the electromagnetic sound sensor comprises an electromagnet and a coil, the coil is arranged in a fixedly-arranged magnetic field of the electromagnet, the coil is connected with the excitation end and the signal receiving end of an electromagnetic control unit, and the electromagnet enables the robot to be attracted to a storage tank. The robot has the advantages that the electromagnet is not only used for manufacturing the electromagnetic acoustic sensor, the magnitude of excitation magnetic current is controlled through the electromagnetic control unit, the storage tank is detected through electromagnetic acoustic excitation, and in addition, the robot is attracted to the storage tank through the electromagnet.

The invention relates to a nondestructive detection device and a nondestructive detection method for a barrel-shaped complex curved surface workpiece, belongs to the field of ray nondestructive testing, and solves the problem that a conventional detection device and method cannot realize vertical transillumination detection on any position of the surface of the workpiece and are not suitable for complex curved surface detection. The surface of the workpiece is divided into a plurality of areas, the position and the direction of the center point of each area are determined, the positions and the directions of a detector and a radiation source are determined according to the positions and the directions of the center points, and full-coverage low-redundancy single-wall vertical transillumination detection on the large-size cylindrical complex curved surface workpiece is achieved. The device and method can be widely applied to the field of ray nondestructive detection.

The invention discloses a photoelectric sensor for the safety of a mechanical operation. The system comprises an annular detection head, a control box and system hardware. The annular detection head is connected with the control box trough cables, and the annular detection head comprises an annular housing, a transmitting lens, a receiving lens, an optical filter and a photoelectric emission detection circuit board. According to the photoelectric sensor for the safety of the mechanical operation, the appearance design and the mechanical structure are fitted, the outline is compact, and the occupied space is small; the sensor is simple to install, and can meet different types of punching machineries; no dead zones are in a detection range, so that a full coverage detection of the dangerousarea can be realized; when a manual operation is needed in punching small size material pieces, the difference between the material pieces of most materials and colors and a finger can be automatically recognized, thereby ensuring that a protection device does not interfere with an normal operation; and the setting function is flexible, so that settings of a threshold value and a multi-output modecan be set according to requirements, and different applications can be met.

The invention discloses a full-automatic workpiece surface quality detection device. A workpiece fixing assembly is arranged on a base, a workpiece upper surface detection assembly is arranged on the upper surface of a to-be-detected workpiece, and a workpiece lower surface detection assembly is arranged on the lower surface of the to-be-detected workpiece. The workpiece upper surface detection assembly and the workpiece lower surface detection assembly are both driven by a detection assembly driving assembly to achieve reciprocating motion of the workpiece upper surface detection assembly and the workpiece lower surface detection assembly, and therefore the whole surfaces of the to-be-detected workpiece can be detected in a full coverage mode. The upper surface and the lower surface of the workpiece are detected in an omni-bearing mode through an upper computer, the detection rate is high, back-and-forth motion of a rack can be achieved only by driving a gear to rotate by a positive and reverse rotating motor, and an upper surface quality detection head and a lower surface quality detection head can detect the upper surface and the lower surface of the workpiece in an omni-bearing mode; detection efficiency is greatly improved, and the accuracy of surface quality detection is improved.

The invention discloses an underwater multi-parameter automatic detection device for an underground diaphragm wall. The device comprises a shore-based control unit, an underwater control unit and an underwater execution device. The underwater control unit is electrically connected with the shore-based control unit and the underwater execution device respectively. The device is wide in single detection coverage and large in area, can realize full-coverage detection of underwater structures of a wharf, replaces an original laborious and troublesome manual random point selection inspection working mode, and is high in work efficiency, low in cost and small in work danger; meanwhile, the underwater execution device is large in overall structure, and the underwater execution device is set to bein a frame structure, thereby making the entire device have higher underwater surge resistance capacity; and besides, through arrangement of multiple groups of anti-collision components, collision between the underwater execution device and the wharf wall can be prevented, thereby improving practicality.

The invention relates to a weld defect detection method and device, electronic equipment and a storage medium. The method comprises the steps of determining a first vibration signal corresponding to a target sensor based on a detection signal of the target sensor; determining a first vibration signal corresponding to each non-target sensor based on the detection signal of each non-target sensor; based on the first vibration signals corresponding to the non-target sensors, determining second vibration signals of reference waves corresponding to the non-target sensors at target sensor setting positions; superposing the first vibration signal corresponding to the target sensor and the second vibration signal of each reference wave at the setting position of the target sensor to obtain a superposed third vibration signal; and determining weld defect data in the to-be-detected component based on the superposed third vibration signal. The device can be used for detecting local defects in a long welding seam structure with high transverse resolution capability and high axial positioning precision, and can realize remote in-service quality monitoring.

The invention discloses a device and a method for detecting the integrity of a paint film of an enameled wire. The device comprises a test circuit (11) and a detection container (12) containing a non-toxic liquid conductive medium (13), the non-toxic liquid conductive medium (13) is used for immersing the to-be-tested enameled wire (15); and the test circuit (11) is used for applying an excitation signal to the to-be-tested enameled wire 15 and detecting whether the excitation signal is generated in the non-toxic liquid conductive medium 13. According to the detection device and method, the enameled wire to be detected can be immersed in the non-toxic liquid conducting medium, and whether the paint film is complete or not is determined by detecting whether the excitation signal is generated in the non-toxic liquid conducting medium or not, so that at least one defect existing in a paint film detection scheme in the related technology can be overcome.

The invention relates to the technical field of gas cylinder type structure pressure container detection, and discloses a large-scale cylinder type container and long pipe trailer steel cylinder eddycurrent internal detection device, which comprises: a detection rod, wherein the detection rod comprises a detection rod main body and a plurality of extension rods which are connected with each other; a chuck detachably connected with the cylinder opening of a large-scale cylinder type container or a long pipe trailer steel cylinder; a circumferential rotating mechanism capable of driving the detection rod to rotate in the circumferential direction; an axial driving mechanism for clamping the detection rod and controlling the detection rod to move axially; and a detection mechanism comprisinga rocker arm assembly, wherein the rocker arm assembly is rotatably connected to one side of the detection rod, a detection probe is installed on the rocker arm assembly, a spring telescopic assemblyis arranged on the detection rod, and the spring telescopic assembly is hinged to the rocker arm assembly. According to the invention, the detection difficulty in equipment is broken through, the detection precision is high, the defect omission ratio is low, operation is simple, measurement time consumption is low, and detection efficiency is high.

The invention discloses a nondestructive testing method for adapting to a suspension arm through automatic telescopic obstacle avoidance. The nondestructive testing method comprises a mobile vehicle, a square main body frame mounted on the mobile vehicle, and a top detection assembly, two side detection assemblies and a bottom detection assembly which are mounted on the main body frame, each detection assembly comprises one or more sensor boxes stretching towards the center position of the main body frame, guide wheels attached to the suspension arm are arranged at the bottoms of the sensor boxes, sensors are arranged in the sensor boxes, and the sensor boxes are connected to the main body frame through telescopic structures; comprising the following steps: S1, sleeving the initial position of a suspension arm with a detection assembly on a mobile vehicle; s2, driving the mobile vehicle to move along the extension direction of the suspension arm; s3, reducing the speed of the mobile vehicle through the deformation position of the suspension arm, and enabling the sensor box to pass through the deformation position after the sensor box slowly stretches out and draws back; and S4, continuously driving the mobile vehicle, and collecting detection data through a sensor until the detection is completed. According to the invention, through the deformation position of the suspension arm, the surface defect of the suspension arm can be fully automatically, rapidly and efficiently detected.