Current detection method and system based on LED fault test power supply

Abstract

The present invention provides a current detection method and system based on an LED fault test power supply, wherein the method comprises: step 1: obtaining the current detection requirement of the LED fault test power supply; step 2: determining the driving strategy of the LED fault test power supply based on the deep learning technology and the current detection requirement; step 3: after driving, measuring the voltage value at both ends of the standard sampling resistor; step 4: calculating the loop current value based on the voltage value at both ends, and judging whether the loop current value meets the current detection standard. The current detection method and system based on the LED fault test power supply of the present invention introduces deep learning technology, determines the driving strategy of the LED fault test power supply based on the current detection requirement, and improves the suitability of the driving; introduces a 10mΩ resistor with copper as the main body as the standard sampling resistor, calculates the loop current value based on Ohm's law, and judges whether the loop current value meets the current detection standard, and improves the convenience of current detection.

Description

Technical Field

[0001] The present invention relates to the technical field of current detection, and in particular to a current detection method and system based on an LED fault test power supply. Background Art

[0002] In the detection test of LED lamp controller, the fault state test is an extremely critical test. In the IEC standard and GB / T standard, there are clear requirements for the fault state test, and the short-circuit current capacity of the test circuit should be 160A (-0+10%). In the test circuit, a standard sampling resistor is introduced as a test device, and the voltage across the two ends of the monitoring standard sampling resistor is introduced into the circuit as a reference quantity to realize current monitoring.

[0003] At present, the sampling resistors used for current detection are mainly aluminum shell resistors. Aluminum shell resistors are usually made of internal resistor materials and filling insulating materials (such as quartz sand, ceramics and cement, etc.) and then packaged in an aluminum shell. They are bulky, not conducive to saving space and reducing weight, and have poor detection convenience.

[0004] In view of this, there is an urgent need for a current detection method and system based on an LED fault test power supply to at least solve the above-mentioned deficiencies. Summary of the invention

[0005] One of the purposes of the present invention is to provide a current detection method and system based on an LED fault test power supply, introduce deep learning technology, and determine the driving strategy of the LED fault test power supply according to the current detection requirements, thereby improving the suitability of the driving; introduce a 10mΩ resistor with copper as the main body as a standard sampling resistor, calculate the loop current value based on Ohm's law, and judge whether the loop current value meets the current detection standard, thereby improving the convenience of current detection.

[0006] The current detection method based on the LED fault test power supply provided by the embodiment of the present invention includes:

[0007] Step 1: Obtain the current detection requirements of the LED fault test power supply;

[0008] Step 2: Determine the driving strategy of the LED fault test power supply based on deep learning technology and current detection requirements;

[0009] Step 3: After driving the LED fault test power supply based on the driving strategy, measure the voltage value across the standard sampling resistor;

[0010] Step 4: Calculate the loop current value based on the voltage values ​​at both ends, and determine whether the loop current value meets the current detection standard.

[0011] Preferably, step 1: obtaining the current detection requirement of the LED fault test power supply includes:

[0012] Obtaining the fault status test standard;

[0013] According to the fault state test standard, obtain the short-circuit current capacity of the test circuit;

[0014] Based on the short-circuit current capability, determine the current sensing requirements.

[0015] Preferably, step 2: determining a driving strategy of an LED fault test power supply based on deep learning technology and current detection requirements, including:

[0016] Get the LED power supply parameters of the LED fault test power supply;

[0017] Based on deep learning technology, the test circuit is designed according to the current detection requirements and LED power supply parameters;

[0018] Determine the driving strategy based on the current detection requirements and test circuit.

[0019] Preferably, based on deep learning technology, according to current detection requirements and LED power supply parameters, a test circuit is designed, including:

[0020] Obtain the target circuit design record based on current detection of LED fault test power supply;

[0021] Parsing the target circuit design record to obtain a first circuit design parameter of a first circuit design parameter type;

[0022] Analyze the current detection requirement and obtain a second circuit design parameter of a second circuit design parameter type;

[0023] Matching the second circuit design parameter type with the first circuit design parameter type, and if the match is satisfactory, calculating the parameter similarity value of the corresponding second circuit design parameter and the first circuit design parameter;

[0024] If for each second circuit design parameter, a first circuit design parameter whose corresponding parameter similarity value is greater than or equal to a preset parameter similarity value threshold can be found, the corresponding target circuit design record is used as a training record;

[0025] Based on deep learning technology, the current detection circuit design model is trained according to the training records;

[0026] Based on the current detection circuit design model, the test circuit is designed according to the current detection requirements and LED power supply parameters.

[0027] Preferably, obtaining a target circuit design record based on current detection of an LED fault test power supply includes:

[0028] Obtain multimodal data of historical detection scenarios based on current detection of LED fault test power supplies;

[0029] Determining modal data types of multimodal data;

[0030] Determine the acquisition strategy of the target circuit design record according to different modal data types;

[0031] Based on the acquisition strategy, a target circuit design record is acquired.

[0032] Preferably, according to different types of modal data, determining the acquisition strategy of the target circuit design record includes:

[0033] When the modal data type is on-site video data, the on-site drawing personnel are determined based on the on-site video data and the drawing personnel;

[0034] After the on-site draftsman is determined, a video of the on-site draftsman drawing the target is obtained;

[0035] Based on the image analysis technology and the drawn target video, determine the acquisition strategy of the target circuit design record;

[0036] When the modal data type is on-site voice data, a strategy for acquiring the target circuit design record is determined based on the on-site voice data based on semantic recognition technology.

[0037] Preferably, when the modal data type is on-site voice data, determining an acquisition strategy for the target circuit design record based on the on-site voice data based on semantic recognition technology includes:

[0038] Expanding the on-site speech data on a time axis to obtain a plurality of first recognition semantic items;

[0039] Obtaining a first semantic type of a first identified semantic item;

[0040] Matching the first semantic type with the preset circuit element description semantic type, and if the match is satisfactory, using the corresponding first identification semantic item as the second identification semantic item;

[0041] Traversing each second identification semantic item in turn, and taking the second identification semantic item being traversed as the third identification semantic item;

[0042] Acquire a second identification semantic item adjacent to the third identification semantic item on the time axis and use it as the fourth identification semantic item;

[0043] Determine the semantics of the component connection relationship according to the first identification semantic item between the third identification semantic item and the fourth identification semantic item;

[0044] A connection template corresponding to the circuit element according to the third identification semantic item;

[0045] The semantics and connection templates are determined according to the component connection relationship, and the acquisition strategy of the target circuit design record is determined.

[0046] Preferably, determining the semantics of the component connection relationship according to the first identification semantic item between the third identification semantic item and the fourth identification semantic item comprises:

[0047] attempting to identify a modified semantic term in the first identification semantic term between the third identification semantic term and the fourth identification semantic term;

[0048] If the acquisition attempt is successful, obtain the adjacent interval where the modified semantic item is located;

[0049] In the adjacent interval where the revised semantic item is located, if the adjacent interval is on the left side of the third identified semantic item, the first identified semantic item between the revised semantic item and the third identified semantic item is used as the component connection relationship to determine the semantics; if the adjacent interval is on the right side of the third identified semantic item, the first identified semantic item after the revised semantic item in the corresponding adjacent interval is used as the component connection relationship to determine the semantics;

[0050] In an adjacent interval not including a modified semantic term, the first identification semantic term between the third identification semantic term and the fourth identification semantic term is used as a component connection relationship to determine semantics.

[0051] Preferably, based on the acquisition strategy, acquiring the target circuit design record includes:

[0052] Based on the acquisition strategy, acquire a first pre-selected circuit design record for the same detection scenario;

[0053] traversing the first pre-selected circuit design records in sequence, taking the first pre-selected circuit design record being traversed as the second pre-selected circuit design record, and at the same time taking the remaining first pre-selected circuit design records as the third pre-selected circuit design record;

[0054] determining readable design similarities between the second pre-selected circuit design record and the third pre-selected circuit design record;

[0055] If there are more than one identical readable design similarity items whose design similarity values ​​are less than a preset design similarity value threshold, the corresponding second pre-selected circuit design record is removed;

[0056] After the traversal of the first pre-selected circuit design records is completed, a target circuit design record is determined according to the remaining first pre-selected circuit design records.

[0057] The current detection system based on the LED fault test power supply provided by the embodiment of the present invention includes:

[0058] The current detection requirement acquisition subsystem is used to obtain the current detection requirement of the LED fault test power supply;

[0059] The driving strategy determination subsystem is used to determine the driving strategy of the LED fault test power supply based on deep learning technology and current detection requirements;

[0060] The voltage measurement subsystem is used to measure the voltage value at both ends of the standard sampling resistor after driving the LED fault test power supply based on the driving strategy;

[0061] The standard determination subsystem is used to calculate the loop current value according to the voltage values ​​at both ends and to determine whether the loop current value meets the current detection standard.

[0062] Preferably, the current detection demand acquisition subsystem includes:

[0063] A fault state test standard acquisition module is used to acquire the fault state test standard;

[0064] A short-circuit current capability acquisition module is used to acquire the short-circuit current capability of the test circuit according to the fault state test standard;

[0065] The current detection requirement determination module is used to determine the current detection requirement based on the short-circuit current capability.

[0066] The beneficial effects of the present invention are:

[0067] The present invention introduces deep learning technology and determines the driving strategy of the LED fault test power supply according to the current detection requirements, thereby improving the suitability of the driving; introduces a 10mΩ resistor with copper as the main body as a standard sampling resistor, calculates the loop current value based on Ohm's law, and judges whether the loop current value meets the current detection standard, thereby improving the convenience of current detection.

[0068] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures specifically pointed out in the present application documents.

[0069] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0071] Figure 1 Schematic diagram of a current detection method based on an LED fault test power supply in an embodiment of the present invention;

[0072] Figure 2Schematic diagram of a current detection circuit based on an LED fault test power supply in an embodiment of the present invention;

[0073] Figure 3 Schematic diagram of a current detection system based on an LED fault test power supply in an embodiment of the present invention. DETAILED DESCRIPTION

[0074] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0075] The embodiment of the present invention provides a current detection method based on an LED fault test power supply, such as Figure 1 As shown, including:

[0076] Step 1: Obtain the current detection requirement of the LED fault test power supply; wherein, the LED fault test power supply is: a power supply system used to test the LED in a fault state; the current detection requirement is: 160A (-0+10%);

[0077] Step 2: Based on the deep learning technology and according to the current detection requirements, determine the driving strategy of the LED fault test power supply; wherein the driving strategy is: a strategy for driving the LED fault test power supply to start up to perform corresponding detection of the current detection requirements, such as: how many volts the voltage source is;

[0078] Step 3: After driving the LED fault test power supply based on the driving strategy, measure the voltage value at both ends of the standard sampling resistor; wherein the standard sampling resistor is: a 10mΩ resistor with copper as the main body;

[0079] Step 4: Calculate the loop current value based on the voltage values ​​at both ends, and determine whether the loop current value meets the current detection standard. The loop current value is calculated based on Ohm's law; determining whether the loop current value meets the current detection standard refers to determining whether the short-circuit current capacity of the test loop reaches 160A (-0+10%).

[0080] The working principle and beneficial effects of the above technical solution are:

[0081] In the detection circuit, a 10mΩ resistor is introduced as a standard sampling resistor. By monitoring the voltage across the 10mΩ resistor, it is introduced into the circuit as a reference quantity to achieve current monitoring. Figure 2 As shown, where T is a 10mΩ resistor; is the power supply voltage; DUT is the sample under test; R is the additional wire or resistor for current regulation; X1, X2, X3, X4 are the terminals of the additional wire or resistor; A and B are the terminals for short circuit and connection of control device; CP is the current detector.

[0082] The 10mΩ resistor product is made of copper-based metal materials with a specific resistivity coefficient. Taking into full consideration the influence of temperature on the stability of the resistance value under the condition of a large current of 160A in the test circuit, a specific structure of 360mm long, 25mm wide and 2mm thick is designed, and the mounting components are screws, washers and nuts. In actual product applications, the 10mΩ resistor is connected in series to the test circuit. By measuring the voltage V across the 10mΩ resistor, the current I flowing through the 10mΩ resistor can be calculated by Ohm's law, thereby accurately evaluating whether the current value of the circuit meets the standard requirements (160A+10%).

[0083] The present invention introduces deep learning technology and determines the driving strategy of the LED fault test power supply according to the current detection requirements, thereby improving the suitability of the driving; introduces a 10mΩ resistor with copper as the main body as a standard sampling resistor, calculates the loop current value based on Ohm's law, and judges whether the loop current value meets the current detection standard, thereby improving the convenience of current detection.

[0084] In one embodiment, step 1: obtaining the current detection requirement of the LED fault test power supply includes:

[0085] Obtain the fault state test standard; wherein the fault state test standard is: EC standard and GB / T standard;

[0086] According to the fault state test standard, the short-circuit current capacity of the test circuit is obtained; among which, the short-circuit current capacity is: 160A (-0+10%);

[0087] Based on the short-circuit current capability, determine the current sensing requirements.

[0088] The working principle and beneficial effects of the above technical solution are:

[0089] The present invention introduces a fault state test standard to obtain the short-circuit current capacity of the test loop, thereby improving the rationality of the current detection requirement determination process.

[0090] In one embodiment, step 2: determining a driving strategy of an LED fault test power supply based on deep learning technology and current detection requirements includes:

[0091] Obtaining LED power supply parameters of the LED fault test power supply; wherein the LED power supply parameters are: electrical characteristics required for the LED fault test power supply to work normally;

[0092] Based on deep learning technology, the test circuit is designed according to the current detection requirements and LED power supply parameters; the test circuit is: the current detection circuit designed based on deep learning technology;

[0093] According to the current detection requirements and the test circuit, the driving strategy is determined. The driving strategy is: a strategy for controlling the LED power supply.

[0094] The working principle and beneficial effects of the above technical solution are:

[0095] The present invention introduces deep learning technology and designs a test circuit according to current detection requirements and LED power supply parameters, thereby improving the suitability of the detection circuit design; and performs corresponding driving according to the current detection requirements and the test circuit, thereby improving the rationality of the driving.

[0096] In one embodiment, based on deep learning technology, a test loop is designed according to current detection requirements and LED power supply parameters, including:

[0097] Obtain a target circuit design record based on current detection of an LED fault test power supply; wherein the target circuit design record is: a circuit design scheme and parameters recorded when designing a current detection circuit of an LED fault test power supply, including a circuit diagram, a component list, and a design intent, etc.;

[0098] Parse the target circuit design record to obtain a first circuit design parameter of a first circuit design parameter type; wherein the circuit design parameter type is: circuit element type; the circuit design parameter is: element parameter value, such as: how many ohms the resistance is;

[0099] Analyze the current detection requirement and obtain a second circuit design parameter of a second circuit design parameter type;

[0100] Matching the second circuit design parameter type with the first circuit design parameter type, and if the match is satisfactory, calculating the parameter similarity value of the corresponding second circuit design parameter and the first circuit design parameter; wherein the parameter similarity value is: the parameter matching degree of circuit design parameters of the same circuit design parameter type;

[0101] If each second circuit design parameter can find a first circuit design parameter whose corresponding parameter similarity value is greater than or equal to a preset parameter similarity value threshold, the corresponding target circuit design record is used as a training record; wherein the parameter similarity value threshold is manually preset, such as: 90;

[0102] Based on deep learning technology, the current detection circuit design model is trained according to the training records; wherein the current detection circuit design model is: an AI model that replaces manual design of the corresponding current detection circuit according to the current detection requirements;

[0103] Based on the current detection circuit design model, the test circuit is designed according to the current detection requirements and LED power supply parameters.

[0104] The working principle and beneficial effects of the above technical solution are:

[0105] The present invention introduces the historical target circuit design record of current detection based on LED fault test power supply, and obtains the first circuit design parameter of the first circuit design parameter type. According to the local current detection requirements, obtain the second circuit design parameter of the second circuit design parameter type; determine the parameter similarity value of the second circuit design parameter and the first circuit design parameter when the circuit design parameter type matches, and screen each second circuit design parameter to find the target circuit design record of the first circuit design parameter whose corresponding parameter similarity value is greater than or equal to the preset parameter similarity value threshold as a training record; based on the deep learning technology, train the current detection circuit design model according to the training record; the current detection circuit design model designs a suitable test circuit according to the current detection requirements and LED power supply parameters, thereby improving the rationality of the test circuit design.

[0106] In one embodiment, obtaining a target circuit design record based on current detection of an LED fault test power supply includes:

[0107] Acquire multimodal data of the detection scenarios of the current detection based on the LED fault test power supply in history; wherein the detection scenario is: the environment of the current detection based on the LED fault test power supply in history; the multimodal data is: a combination of multiple different types of data, such as: numerical data, text data, image data and sound data;

[0108] Determine the modal data type of the multimodal data; wherein the modal data type is: Data type of the multimodal data:

[0109] Determine the acquisition strategy of the target circuit design record according to different modal data types; wherein the acquisition strategy is: a method for collecting and selecting the target circuit design record according to different modal data types;

[0110] Based on the acquisition strategy, a target circuit design record is acquired.

[0111] The working principle and beneficial effects of the above technical solution are:

[0112] The present invention introduces multimodal data of a detection scene, and determines a corresponding acquisition strategy for a target circuit design record according to different modal data types of the multimodal data; the target circuit design record is acquired in a targeted manner based on the acquisition strategy, thereby improving the acquisition accuracy of the target circuit design record.

[0113] In one embodiment, according to different modal data types, determining an acquisition strategy for a target circuit design record includes:

[0114] When the modal data type is on-site video data, the on-site draftsman is determined based on the on-site video data and the features determined by the draftsman; the on-site video data is: a video of an actual test scene recorded when the current detection is performed based on the LED fault test power supply in the past; the draftsman is determined by: holding a pen, using circuit drawing software, etc.;

[0115] After the on-site draftsman is determined, a drawing target video of the on-site draftsman is obtained; wherein the drawing target video is: a local video data containing the on-site draftsman;

[0116] Based on the image analysis technology, according to the drawing target video, determine the acquisition strategy of the target circuit design record; wherein, based on the image analysis technology, according to the drawing target video, determine the acquisition strategy of the target circuit design record: based on the image analysis technology, analyze the drawing completion image in the drawing target video, and extract the strategy of the target circuit design from the drawing completion image;

[0117] When the modal data type is on-site voice data, the acquisition strategy of the target circuit design record is determined based on the semantic recognition technology and the on-site voice data. The on-site voice data is the on-site voice recorded during the current detection based on the LED fault test power supply in history; the acquisition strategy of the target circuit design record is determined based on the semantic recognition technology and the on-site voice data: the semantics of the on-site voice data is recognized based on the semantic recognition technology, and the strategy of the target circuit design is determined based on the semantic understanding result.

[0118] The working principle and beneficial effects of the above technical solution are:

[0119] The present invention introduces two acquisition strategies for extracting target circuit design records corresponding to two different modal data. The first one is that when the modal data type is on-site video data, features are determined according to the on-site video data and the draftsman, a drawing target video containing the on-site draftsman is determined, and the drawing completion image in the drawing target video is analyzed based on image analysis technology, and the target circuit design is extracted from the drawing completion image; the second one is that when the modal data type is on-site voice data, the target circuit design is determined according to the semantic recognition result of the on-site voice data, thereby improving the rationality of acquiring the target circuit design record.

[0120] In one embodiment, when the modal data type is on-site voice data, determining an acquisition strategy for a target circuit design record based on the on-site voice data based on semantic recognition technology includes:

[0121] Expanding the on-site speech data on a time axis to obtain a plurality of first identifiable semantic items; wherein the first identifiable semantic items are: the results of analyzing the on-site speech data based on a semantic analysis technology, and the first identifiable semantic items are expanded on the time axis according to the time position of semantic generation;

[0122] Acquire a first semantic type of the first identified semantic item; wherein the first semantic type is: a description type of the first identified semantic item, such as: a connection relationship, or: a preposition;

[0123] The first semantic type is matched with the preset circuit element description semantic type. If the match is satisfied, the corresponding first identification semantic item is used as the second identification semantic item; wherein the preset circuit element description semantic type is: a type of description semantics used to describe the circuit element; the second identification semantic item is: the circuit element description semantics in the first identification semantic item, such as: "10mΩ resistor", "power supply voltage", etc.;

[0124] Traversing each second identification semantic item in turn, and taking the second identification semantic item being traversed as the third identification semantic item;

[0125] Acquire a second identification semantic item adjacent to the third identification semantic item on the time axis and use it as the fourth identification semantic item;

[0126] Determine the semantics of the component connection relationship according to the first identification semantic item between the third identification semantic item and the fourth identification semantic item;

[0127] A connection template for a circuit element corresponding to the third identification semantic item; wherein the connection template is a template for generating a connection relationship of the circuit element by comparing the connection relationship description semantics of the circuit element corresponding to the third identification semantic item, such as: which element is connected to the left terminal and which element is connected to the right terminal;

[0128] The semantics and connection templates are determined according to the component connection relationship, and the acquisition strategy of the target circuit design record is determined.

[0129] The working principle and beneficial effects of the above technical solution are:

[0130] The present invention expands the first identification semantic item corresponding to the on-site voice data on the time axis to obtain multiple first identification semantic items; matches the first semantic type of the first identification semantic item with the circuit element description semantic type; determines the circuit element description semantics (second identification semantic item) in the first identification semantic item; traverses the second identification semantic items in sequence, determines the third identification semantic item being traversed, and obtains the fourth identification semantic item adjacent to the third identification semantic item on the time axis. The third identification semantic item and the fourth identification semantic item are both circuit element description semantics. The first identification semantic item between the two generally contains the description semantics of the connection relationship between the two in the design circuit. The connection template of the circuit element corresponding to the third identification semantic item is introduced, the semantics and the connection template are determined according to the element connection relationship, and the acquisition strategy of the target circuit design record is determined, which avoids the problem of inaccurate circuit design obtained based on the semantic analysis result and is more reasonable.

[0131] In one embodiment, determining the semantics of the component connection relationship according to the first identification semantic item between the third identification semantic item and the fourth identification semantic item includes:

[0132] Attempt to identify a modified semantic term in the first identification semantic term between the third identification semantic term and the fourth identification semantic term; wherein the modified semantic term is: a semantic term indicating negation, and the semantic term indicating negation overrules the first identification semantic term preceding it by default;

[0133] If the acquisition attempt is successful, the adjacent interval where the modified semantic item is located is acquired; wherein the adjacent interval is: the semantic interval between the third identification semantic item and the fourth identification semantic item;

[0134] In the adjacent interval where the revised semantic item is located, if the adjacent interval is on the left side of the third identified semantic item, the first identified semantic item between the revised semantic item and the third identified semantic item is used as the component connection relationship to determine the semantics; if the adjacent interval is on the right side of the third identified semantic item, the first identified semantic item after the revised semantic item in the corresponding adjacent interval is used as the component connection relationship to determine the semantics;

[0135] In an adjacent interval not including a modified semantic term, the first identification semantic term between the third identification semantic term and the fourth identification semantic term is used as a component connection relationship to determine semantics.

[0136] The working principle and beneficial effects of the above technical solution are:

[0137] The present invention determines a modified semantic item between a third identification semantic item and a fourth identification semantic item. In an adjacent interval containing the modified semantic item, if the adjacent interval is on the left side of the third identification semantic item, the first identification semantic item between the modified semantic item and the third identification semantic item is used as the component connection relationship determination semantics. If the adjacent interval is on the right side of the third identification semantic item, the first identification semantic item after the modified semantic item in the corresponding adjacent interval is used as the component connection relationship determination semantics. In an adjacent interval not containing the modified semantic item, the first identification semantic item between the third identification semantic item and the fourth identification semantic item is directly used as the component connection relationship determination semantics. The modified semantic item is determined in each adjacent interval of the third identification semantic item respectively, and the actual component connection relationship determination semantics is determined based on the modified semantic item, thereby improving the acquisition accuracy of the component connection relationship determination semantics.

[0138] In one embodiment, based on the acquisition strategy, acquiring the target circuit design record includes:

[0139] Based on the acquisition strategy, a first pre-selected circuit design record of the same detection scenario is acquired; wherein the first pre-selected circuit design record is: a circuit design record acquired based on different acquisition strategies in the same detection scenario, such as: a circuit design acquired based on semantic analysis, and another example: a circuit design acquired based on image recognition;

[0140] traversing the first pre-selected circuit design records in sequence, taking the first pre-selected circuit design record being traversed as the second pre-selected circuit design record, and at the same time taking the remaining first pre-selected circuit design records as the third pre-selected circuit design record;

[0141] Determining readable design similarities between the second pre-selected circuit design record and the third pre-selected circuit design record; wherein the readable design similarities are: similar design elements or features that can be identified and compared in circuit design, such as: circuit layout, component usage, and signal path;

[0142] If there are more than one identical readable design similarity items whose design similarity values ​​are less than a preset design similarity value threshold, the corresponding second pre-selected circuit design record is removed; wherein the design similarity value is the similarity degree of the readable design similarity items; the preset design similarity value threshold is: 0.9;

[0143] After the traversal of the first pre-selected circuit design records is completed, a target circuit design record is determined according to the remaining first pre-selected circuit design records.

[0144] The working principle and beneficial effects of the above technical solution are:

[0145] The present invention obtains different first pre-selected circuit design records in the same detection scene based on different acquisition strategies. Since the information obtained by parsing different modal data is different, the extracted first pre-selected circuit design records may also be different. For example, voice data can clearly obtain component composition, but the connection relationship between components can only be determined partially based on semantic descriptions, while video data can obtain clear connection relationships, but some circuit designs may have unclear identification problems. However, different first pre-selected circuit design records are extracted from the same detection scene, so the corresponding circuit designs are also consistent. Therefore, the first pre-selected circuit design records are traversed in sequence, and the first pre-selected circuit design records being traversed are The first pre-selected circuit design record is used as the second pre-selected circuit design record, and the remaining first pre-selected circuit design records are used as the third pre-selected circuit design records; the readable design similarities of the second pre-selected circuit design record and the third pre-selected circuit design record are determined, and if there are more than one identical readable design similarities whose design similarity values ​​are less than a preset design similarity value threshold, it means that the corresponding second pre-selected circuit design record and the third pre-selected circuit design record extracted from two homologous but different modal data have consistent readable design similarities, then the corresponding second pre-selected circuit design record is unreliable and is eliminated, and then the target circuit design record is obtained by fusion based on the remaining first pre-selected circuit design records, thereby further improving the accuracy of obtaining the target circuit design record.

[0146] The embodiment of the present invention provides a current detection system based on an LED fault test power supply, such as Figure 3 As shown, including:

[0147] The current detection requirement acquisition subsystem 1 is used to obtain the current detection requirement of the LED fault test power supply;

[0148] A driving strategy determination subsystem 2 is used to determine the driving strategy of the LED fault test power supply based on deep learning technology and current detection requirements;

[0149] The voltage measurement subsystem 3 is used to measure the voltage value at both ends of the standard sampling resistor after driving the LED fault test power supply based on the driving strategy;

[0150] The standard determination subsystem 4 is used to calculate the loop current value according to the voltage values ​​at both ends, and to determine whether the loop current value meets the current detection standard.

[0151] In one embodiment, the current detection requirement acquisition subsystem includes:

[0152] A fault state test standard acquisition module is used to acquire the fault state test standard;

[0153] A short-circuit current capability acquisition module is used to acquire the short-circuit current capability of the test circuit according to the fault state test standard;

[0154] The current detection requirement determination module is used to determine the current detection requirement based on the short-circuit current capability.

[0155] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A current detection method based on an LED fault test power supply, characterized in that: include: Step 1: Obtain the fault state test standard, obtain the short-circuit current capacity of the test circuit according to the fault state test standard, and determine the current detection requirement based on the short-circuit current capacity; Step 2: Determine the driving strategy of the LED fault test power supply based on deep learning technology and current detection requirements; Among them, step 2: based on deep learning technology and according to current detection requirements, determine the driving strategy of the LED fault test power supply, including: Get the LED power supply parameters of the LED fault test power supply; Based on deep learning technology, the test circuit is designed according to the current detection requirements and LED power supply parameters; Determine the driving strategy based on current detection requirements and test circuits; Step 3: After driving the LED fault test power supply based on the driving strategy, measure the voltage value across the standard sampling resistor; Step 4: Calculate the loop current value according to the voltage values ​​at both ends, and determine whether the loop current value meets the current detection standard; The current detection method is used for detection test of LED lamp controller.

2. The current detection method based on LED fault test power supply according to claim 1, characterized in that: Based on deep learning technology, the test circuit is designed according to the current detection requirements and LED power supply parameters, including: Obtain the target circuit design record based on current detection of LED fault test power supply; Parsing the target circuit design record to obtain a first circuit design parameter of a first circuit design parameter type; Analyze the current detection requirement and obtain a second circuit design parameter of a second circuit design parameter type; Matching the second circuit design parameter type with the first circuit design parameter type, and if the match is satisfactory, calculating the parameter similarity value of the corresponding second circuit design parameter and the first circuit design parameter; If for each second circuit design parameter, a first circuit design parameter whose corresponding parameter similarity value is greater than or equal to a preset parameter similarity value threshold can be found, the corresponding target circuit design record is used as a training record; Based on deep learning technology, the current detection circuit design model is trained according to the training records; Based on the current detection circuit design model, the test circuit is designed according to the current detection requirements and LED power supply parameters.

3. The current detection method based on LED fault test power supply according to claim 2, characterized in that: Get the target circuit design record for current detection based on LED fault test power supply, including: Obtain multimodal data of historical detection scenarios based on current detection of LED fault test power supplies; Determining modal data types of multimodal data; Determine the acquisition strategy of the target circuit design record according to different modal data types; Based on the acquisition strategy, a target circuit design record is acquired.

4. The current detection method based on LED fault test power supply according to claim 3, characterized in that: According to the different modal data types, determine the acquisition strategy of the target circuit design record, including: When the modal data type is on-site video data, the on-site drawing personnel are determined based on the on-site video data and the drawing personnel; After the on-site draftsman is determined, a video of the on-site draftsman drawing the target is obtained; Based on the image analysis technology and the drawn target video, determine the acquisition strategy of the target circuit design record; When the modal data type is on-site voice data, a strategy for acquiring the target circuit design record is determined based on the on-site voice data based on semantic recognition technology.

5. The current detection method based on LED fault test power supply according to claim 4, characterized in that: When the modal data type is on-site voice data, a strategy for acquiring the target circuit design record is determined based on the semantic recognition technology and the on-site voice data, including: Expanding the on-site speech data on a time axis to obtain a plurality of first recognition semantic items; Obtaining a first semantic type of a first identified semantic item; Matching the first semantic type with the preset circuit element description semantic type, and if the match is satisfactory, using the corresponding first identification semantic item as the second identification semantic item; Traversing each second identification semantic item in turn, and taking the second identification semantic item being traversed as the third identification semantic item; Acquire a second identification semantic item adjacent to the third identification semantic item on the time axis and use it as the fourth identification semantic item; Determine the semantics of the component connection relationship according to the first identification semantic item between the third identification semantic item and the fourth identification semantic item; A connection template corresponding to the circuit element according to the third identification semantic item; The semantics and connection templates are determined according to the component connection relationship, and the acquisition strategy of the target circuit design record is determined.

6. The current detection method based on LED fault test power supply according to claim 5, characterized in that: Determining the semantics of the component connection relationship according to the first identification semantic item between the third identification semantic item and the fourth identification semantic item includes: attempting to identify a modified semantic term in the first identification semantic term between the third identification semantic term and the fourth identification semantic term; If the acquisition attempt is successful, obtain the adjacent interval where the modified semantic item is located; In the adjacent interval where the revised semantic item is located, if the adjacent interval is on the left side of the third identified semantic item, the first identified semantic item between the revised semantic item and the third identified semantic item is used as the component connection relationship to determine the semantics; if the adjacent interval is on the right side of the third identified semantic item, the first identified semantic item after the revised semantic item in the corresponding adjacent interval is used as the component connection relationship to determine the semantics; In an adjacent interval not including a modified semantic term, the first identification semantic term between the third identification semantic term and the fourth identification semantic term is used as a component connection relationship to determine semantics.

7. The current detection system based on LED fault test power supply is characterized in that: include: A current detection requirement acquisition subsystem is used to acquire the fault state test standard, acquire the short-circuit current capability of the test loop according to the fault state test standard, and determine the current detection requirement based on the short-circuit current capability; The driving strategy determination subsystem is used to determine the driving strategy of the LED fault test power supply based on deep learning technology and current detection requirements; Among them, the driving strategy determines the subsystem and performs the following operations: Get the LED power supply parameters of the LED fault test power supply; Based on deep learning technology, the test circuit is designed according to the current detection requirements and LED power supply parameters; Determine the driving strategy based on current detection requirements and test circuits; The voltage measurement subsystem is used to measure the voltage value at both ends of the standard sampling resistor after driving the LED fault test power supply based on the driving strategy; The standard determination subsystem is used to calculate the loop current value according to the voltage value at both ends and determine whether the loop current value meets the current detection standard; The current detection system is used for detection test of LED lamp controller.

Images