A method and system for decomposing synthetic signals

Abstract

The present invention discloses a synthetic signal decomposition method and system, comprising: obtaining a synthetic signal to be decomposed and a basic signal library; inputting the synthetic signal to be decomposed into a trained neural network model for signal decomposition to obtain a decomposed meta-signal; the neural network model is obtained by training the basic signal library. The neural network model is trained by the basic signal library, and the synthetic signal to be decomposed is input into the trained neural network model for signal decomposition, thereby realizing the decomposition of the synthetic signal and reducing the cost.

Description

Technical Field

[0001] The present disclosure relates to a method and system for decomposing a synthetic signal. Background Art

[0002] The statements in this section merely provide background information related to the present disclosure and do not necessarily constitute prior art.

[0003] Signal-oriented test languages ​​include ATLAS and ATML. ATLAS was born in the 1960s. In 1962, the American ARINC company hosted the development of the aviation standard test description language ATLAS to solve the standardization problem of test description language in the aviation field. In 1976, the Institute of Electrical and Electronics Engineers (IEEE) began to take over the standardization of ATLAS. In 1995, the Standard Coordinating Committee SCC20 (Standard Coordinating Committee 20) established by IEEE released the latest standard of ATLAS, IEEE Std 716-1995. ATLAS has advantages such as device independence and strong scalability. However, with the growth of test needs, ATLAS has gradually exposed shortcomings such as increasingly complex language and vague signal definition. For this reason, SCC20 has begun to improve the ATLAS standard, resulting in a new generation of test language---ATLAS 2000, and on this basis, the IEEE 1641 standard was formulated, which is used for signal and test definition. The IEEE 1641 standard solves the problem of test description standardization and instrument interchangeability, but it cannot solve the problem of test program migration. For this reason, IEEE has developed the ATML standard family since 2002, resulting in the IEEE 1671 series of standards, which use XML language to describe test information. This standard family references the IEEE 1641 standard and uses signals to describe the test requirements, instrument capabilities, test descriptions, and other information of the unit under test (UUT).

[0004] BSC is a set of components defined by IEEE 1641, which provides the most basic building blocks for creating signals. BSC defines an extension mechanism. Signals not defined in the most basic signal library can be interconnected by multiple BSC modules according to the standard framework to generate complex signals to meet test requirements and form a standard TSF layer. The process of using BSC to create a complex signal model is as follows: Figure 7 shown.

[0005] When test developers have Figure 7When the right signal model is used, it is easy to separate the complex signal after synthesis into the basic signal before synthesis. However, in most cases, the test developer will only get the signal attribute description of the complex signal after synthesis. In this case, it will be difficult for the tester to carry out test development work based on the attribute description of such a signal, and the signal needs to be decomposed to obtain the basic signal for synthesizing the signal.

[0006] There are currently two methods for decomposing synthetic signals:

[0007] 1. It is necessary to manually reverse decompose the properties of the synthetic signal based on experience. Manual reverse decomposition of the signal is time-consuming and labor-intensive, with high labor costs, low efficiency, and long time consumption. In addition, when the synthetic signal is highly complex, it is often almost impossible to decompose it.

[0008] 2. Software developers are required to develop synthetic signal decomposition software through traditional logic symbol-based programming methods, which is difficult and needs to be redeveloped when the basic signal library is changed, resulting in poor reusability. Summary of the invention

[0009] In order to solve the above problems, the present disclosure proposes a method and system for decomposing synthetic signals. A neural network model is trained through a basic signal library, and the synthetic signal to be decomposed is input into the trained neural network model for signal decomposition, thereby realizing the decomposition of the synthetic signal and reducing the cost.

[0010] To achieve the above objectives, the present disclosure adopts the following technical solutions:

[0011] In one or more embodiments, a method for decomposing a composite signal is proposed, comprising:

[0012] Obtaining the synthetic signal to be decomposed and the basic signal library;

[0013] Input the synthetic signal to be decomposed into the trained neural network model for signal decomposition to obtain the decomposed meta-signal;

[0014] The neural network model is trained by the basic signal library.

[0015] In one or more embodiments, a system for decomposing a composite signal is provided, comprising:

[0016] A basic signal synthesis module selects signals from the basic signal library for synthesis to obtain a new synthesized signal;

[0017] The neural network training module creates a neural network to be trained, trains the neural network according to the synthesized new signal and the signal involved in the synthesis, and obtains a trained neural network model.

[0018] The signal decomposition module inputs the synthetic signal to be decomposed into the trained neural network model to perform signal decomposition.

[0019] In one or more embodiments, a computer storage medium is provided for storing computer instructions, and when the computer instructions are executed by a processor, the steps of the method for decomposing a synthetic signal are completed.

[0020] Compared with the prior art, the present invention has the following beneficial effects:

[0021] 1. The present invention trains a neural network model through a basic signal library, and inputs the synthetic signal to be decomposed into the trained neural network model for signal decomposition, thereby realizing the decomposition of the synthetic signal and reducing the cost. It solves the problem that the existing manual reverse decomposition of signals is time-consuming and labor-intensive, with high labor costs, low efficiency, and long time consumption, and often results in the problem that when the synthetic signal is highly complex, it is almost impossible to decompose it. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The drawings in the specification, which constitute a part of the present application, are used to provide further understanding of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute improper limitations on the present application.

[0023] Figure 1 The overall flow chart of the disclosed method is as follows;

[0024] Figure 2 This is a basic signal synthesis flow chart of the present disclosure;

[0025] Figure 3 This is a training flow chart of the neural network model disclosed in the present invention;

[0026] Figure 4 The figure is a decomposition flow chart of the synthesized signal of the present disclosure.

[0027] Figure 5 is a structural block diagram of the disclosed system;

[0028] Figure 6 The existing hierarchical model of the IEEE 1641 standard;

[0029] Figure 7 The process of creating complex signal models for existing BSCs. DETAILED DESCRIPTION

[0030] The present disclosure is further described below in conjunction with the accompanying drawings and embodiments.

[0031] It should be noted that the following detailed descriptions are illustrative and are intended to provide further explanation of the present application. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present application belongs.

[0032] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0033] In the present disclosure, terms such as "upper", "lower", "left", "right", "front", "back", "vertical", "horizontal", "side", "bottom" and the like indicating directions or positional relationships are based on the directions or positional relationships shown in the accompanying drawings. They are relational words determined only for the convenience of describing the structural relationships of the various parts or elements of the present disclosure, and do not specifically refer to any part or element in the present disclosure and should not be understood as limitations on the present disclosure.

[0034] In the present disclosure, terms such as "fixed connection", "connected", "connection", etc. should be understood in a broad sense, indicating that it can be fixedly connected, integrally connected or detachably connected; it can be directly connected or indirectly connected through an intermediate medium. For relevant scientific research or technical personnel in this field, the specific meanings of the above terms in the present disclosure can be determined according to specific circumstances, and they cannot be understood as limitations on the present disclosure.

[0035] Example 1

[0036] Signal-oriented test languages ​​include ATLAS and ATML. ATLAS was born in the 1960s. In 1962, the American ARINC company hosted the development of the aviation standard test description language ATLAS to solve the standardization problem of test description language in the aviation field. In 1976, the Institute of Electrical and Electronics Engineers (IEEE) began to take over the standardization of ATLAS. In 1995, the Standard Coordinating Committee SCC20 (Standard Coordinating Committee 20) established by IEEE released the latest standard of ATLAS, IEEE Std 716-1995. ATLAS has advantages such as device independence and strong scalability. However, with the growth of test needs, ATLAS has gradually exposed shortcomings such as increasingly complex language and vague signal definition. For this reason, SCC20 has begun to improve the ATLAS standard, resulting in a new generation of test language---ATLAS 2000, and on this basis, the IEEE 1641 standard was formulated, which is used for signal and test definition. The IEEE 1641 standard solves the problem of test description standardization and instrument interchangeability, but it cannot solve the problem of test program migration. For this reason, IEEE has developed the ATML standard family since 2002, resulting in the IEEE 1671 series of standards, which use XML language to describe test information. This standard family references the IEEE 1641 standard and uses signals to describe the test requirements, instrument capabilities, test descriptions, and other information of the unit under test (UUT).

[0037] Each layer is built on the basis of the previous layer. It includes the signal modeling language layer, the basic signal component layer, the test signal framework layer and the test process language layer, such as Figure 6 As shown in the figure. BSC is a set of components defined by IEEE 1641, which provides the most basic building blocks for creating signals. BSC defines an extension mechanism. Signals not defined in the most basic signal library can be interconnected by multiple BSC modules according to the standard framework to generate complex signals to meet test requirements and form a standard TSF layer. The process of using BSC to create a complex signal model is shown in the figure. Figure 7 shown.

[0038] When test developers have Figure 7 When the right signal model is used, it is easy to split the complex signal after synthesis into the basic signal before synthesis. However, in most cases, the test developer will only get the signal attribute description of the complex signal after synthesis. In this case, it will be difficult for the tester to carry out test development work based on the attribute description of this signal. Currently, there are two methods for decomposing synthetic signals:

[0039] 1. It is necessary to manually reverse decompose the properties of the synthetic signal based on experience, which is a time-consuming and labor-intensive process.

[0040] 2. Software developers are required to develop synthetic signal decomposition software through traditional logic symbol-based programming methods, which is difficult and needs to be redeveloped when the basic signal library is changed, resulting in poor reusability.

[0041] Therefore, the present embodiment discloses a method for decomposing a synthetic signal, which is used to decompose a complex synthetic signal into a basic signal. The neural network model is trained through a basic signal library, and the synthetic signal to be decomposed is input into the trained neural network model for signal decomposition, thereby realizing the decomposition of the synthetic signal and reducing the cost. This solves the problem that the existing manual reverse decomposition of the signal is time-consuming and labor-intensive, with high labor costs, low efficiency, and long time consumption, and that the synthetic signal is often almost impossible to decompose when the complexity of the synthetic signal is high.

[0042] It should be noted that the IEEE1641 basic signal library allows signal extension. In addition to signal synthesis, this extension method also allows the creation of new signals. The created signal can describe the signal by itself. The decomposition method of the synthesized signal disclosed in this embodiment cannot decompose the signal described by itself.

[0043] Artificial Neural Network (ANN) is an abstraction of the human brain neural network from the perspective of information processing, establishing a simple model and forming different networks according to different connection methods. A neural network is a computational model composed of a large number of nodes (or neurons) connected to each other.

[0044] Artificial neural network is a nonlinear, adaptive information processing system composed of a large number of interconnected processing units. It is proposed based on the research results of modern neuroscience, and attempts to process information by simulating the way brain neural networks process and memorize information. Artificial neural networks have four basic characteristics:

[0045] (1) Nonlinearity. Nonlinear relationships are a universal feature of nature. Artificial neurons are in two different states: activated or inhibited. This behavior is mathematically expressed as a nonlinear relationship.

[0046] (2) Non-localization. A neural network is usually composed of multiple neurons that are widely connected. The overall behavior of a system depends not only on the characteristics of a single neuron, but may also be mainly determined by the interactions and connections between units. The non-localization of the brain is simulated by a large number of connections between units.

[0047] (3) Very qualitative. Artificial neural networks have the ability to adapt, self-organize, and self-learn. Not only can the information processed by the neural network have various changes, but while processing information, the nonlinear dynamic system itself is also constantly changing. This change process is described by adopting an iterative process.

[0048] (4) Non-convexity. The evolutionary direction of a system will depend on a certain state function under certain conditions. For example, the energy function, whose extreme value corresponds to the relatively stable state of the system. Non-convexity means that this function has multiple extreme values, so the system has multiple relatively stable equilibrium states, which will lead to the diversity of system evolution.

[0049] Artificial neural network is a parallel distributed system that adopts a mechanism completely different from traditional artificial intelligence and information processing technology. It overcomes the shortcomings of traditional artificial intelligence based on logical symbols in processing intuitive and unstructured information, and has the characteristics of self-adaptation, self-organization and real-time learning.

[0050] A method for decomposing a synthetic signal, comprising:

[0051] Obtaining the synthetic signal to be decomposed and the basic signal library;

[0052] Input the synthetic signal to be decomposed into the trained neural network model for signal decomposition to obtain the decomposed meta-signal, which is the basic signal constituting the synthetic signal to be decomposed and is included in the basic signal library;

[0053] The neural network model is trained by the basic signal library.

[0054] Furthermore, it also includes verifying the decomposition results;

[0055] The decomposed meta-signals are resynthesized and compared with the synthesized signal to be decomposed. If they are the same, the verification is passed; if they are different, the verification fails.

[0056] Preferably, if the verification passes, the decomposition result is output;

[0057] If the verification fails, the neural network model is retrained, and the synthetic signal to be decomposed is decomposed by the retrained neural network model until the decomposition result passes the verification.

[0058] Furthermore, the training process of the neural network model is:

[0059] Create the neural network to be trained;

[0060] Get the basic signal library;

[0061] Randomly select basic signals in the basic signal library for synthesis to obtain a new synthesized signal;

[0062] Input all new signals into the neural network for pre-decomposition and verify the pre-decomposition results;

[0063] When the pre-decomposition results of all new signals are correct, the neural network model training is completed.

[0064] Preferably, the process of verifying the pre-decomposition result is:

[0065] The pre-decomposition signal is compared with the basic signal selected for synthesizing the new signal. If they are consistent, the pre-decomposition result is correct. If they are inconsistent, the pre-decomposition result is wrong.

[0066] Preferably, when the pre-decomposition result is wrong, the new signal is pre-decomposed again until the pre-decomposition result is verified to be correct.

[0067] Preferably, the properties of all selected basic signals are combined to perform signal synthesis.

[0068] Preferably, the selected basic signals are the same or different.

[0069] The basic signal library is the IEEE1641-BSC basic signal library, which allows signal extension. In addition to signal synthesis, this extension method also allows the creation of new signals, and the created signal can describe the signal by itself.

[0070] The decomposition method of the synthetic signal is discussed in detail by combining the basic signal synthesis module, neural network training module and signal decomposition module.

[0071] The collaborative relationship among the basic signal synthesis module, neural network training module and signal decomposition module is as follows: Figure 1 As shown, specifically:

[0072] Step 1: Input the IEEE1641-BSC basic signal library into the basic signal synthesis module;

[0073] Step 2: A basic signal synthesis module randomly selects a basic signal from a basic signal library and synthesizes a new signal by a signal synthesis method, wherein the synthesized signal attribute is obtained by taking the union of the attributes of all basic signals involved in the synthesis;

[0074] Step 3: Record the synthesized signals and their properties as well as the basic signals and quantities involved in the synthesis;

[0075] Step 4: Input the basic signals and quantities involved in the synthesis, the synthesized signals and their attributes into the neural network training module;

[0076] Step 5: Set n=0, input the synthesized signal and quantity into the unfinished signal decomposition module obtained through neural network training for pre-decomposition. If the correct result is obtained, n=n+1, and then input the synthesized signal and attribute and the basic signal and quantity involved in the synthesis into the neural network training, establish the network topology relationship, and obtain the neural network model;

[0077] Step 6: After the decomposition is correct m times in a row, m=n, where the size of m is customized by the user, it is considered that the neural network model training has achieved the desired effect, and the training result is output to obtain the signal decomposition module. If m>n, continue the training process in step 2;

[0078] Step 7: Input the composite signal and its attributes to be decomposed and the IEEE1641-BSC basic signal library into the signal decomposition module;

[0079] Step 8: Obtain the element signals and their number sets in the IEEE1641-BSC basic signal library;

[0080] Step 9: Combine the meta-signals obtained after decomposition and verify them with the synthetic signals and attributes before decomposition;

[0081] Step 10: If the verification is successful, the decomposition is correct, and the process continues to step 11; if the verification fails, the decomposition is wrong, and the process continues to step 2;

[0082] Step 11: Output the decomposition results;

[0083] Step 12: End.

[0084] Basic signal synthesis module process, such as Figure 2 As shown, the specific steps are:

[0085] Step 1: Input IEEE1641-BSC basic signal library;

[0086] Step 2: Construct the attribute object of each signal in the basic signal library. You can use the class method to describe the attribute, and the unit of the specific attribute should also be in the description;

[0087] Step 3: Integrate all described objects into a set;

[0088] Step 4: randomly select n signal objects from the set, and these selected signal objects can be repeated but cannot be different from each other;

[0089] Step 5: Combine the signals of the n signal objects by taking the union of the attributes of all signals;

[0090] Step 6: Record the synthesized signal and record it in a synthesized signal set;

[0091] Step 7: record the n signals selected to synthesize the signal and record them in the basic signal set;

[0092] Step 8: End.

[0093] The training process of the neural network model in the neural network training module is as follows Figure 3 As shown:

[0094] Step 1: Create neural network training according to the neural network algorithm;

[0095] Step 2: Input IEEE1641-BSC basic signal library;

[0096] Step 3: Input the synthesized signal set, the number of elements in the signal set is n, n>0;

[0097] Step 4: Initialize i=0;

[0098] Step 5: Determine whether i is greater than n;

[0099] Step 6: If i>n is false, continue to step 7; if i>n is true, jump to step 14;

[0100] Step 7: Select the i-th element in the signal set;

[0101] Step 8: Training through neural network;

[0102] Step nine, obtain a basic signal set through training;

[0103] Step 10: Compare the basic signal set obtained by training with the basic signal set selected for actual synthesis to determine whether they are consistent;

[0104] Step 11: If they are consistent, proceed to step 12; if they are inconsistent, jump to step 13;

[0105] Step 12, i=i+1, jump to step 5;

[0106] Step 13: Adjust the neural training method and jump to step 8;

[0107] Step 14: Output the trained neural network model into a signal decomposition module;

[0108] Step 15: End.

[0109] The process of using the signal decomposition module is as follows Figure 4 As shown, the specific steps are:

[0110] Step 1: Instantiate the signal decomposition module;

[0111] Step 2: Input IEEE1641-BSC basic signal library;

[0112] Step 3: input the synthesized signal to be synthesized;

[0113] Step 4: Obtain the meta-signal set after decomposition;

[0114] Step 5: re-synthesize the obtained element signals;

[0115] Step 6: Determine whether the synthesized signal is the same as the input synthesized signal to be decomposed, if they are the same, proceed to step 7, if they are not the same, jump to step 8;

[0116] Step 7: Output the meta-signal set after decomposition, and jump to step 10;

[0117] Step 8: Prompt decomposition error and request to check the signal decomposition module;

[0118] Step 9: Retrain the neural network;

[0119] Step 10: End.

[0120] A synthetic signal decomposition method proposed in this embodiment is used to solve the decomposition of synthetic signals into basic signals in the signal test development process in the case of an unknown signal model. A neural network model is trained by the basic signal used by the synthetic signal, and the synthetic signal to be decomposed is input into the trained neural network model for decomposition. The extremely difficult task of signal decomposition is solved and converted into training using the characteristics of the neural network to obtain a neuron topology structure for synthetic signal decomposition, which greatly reduces the workload and difficulty of software development and reduces work costs.

[0121] When the BSC basic signal library changes, only the input IEEE1641-BSC basic signal library needs to be modified to achieve the reuse of the entire code without redeveloping the internal logic, which greatly improves the reusability of the device.

[0122] Example 2

[0123] In this embodiment, a synthetic signal decomposition system is disclosed, the structure of which is as follows: Figure 5 As shown in the figure, the signal decomposition module marked on the right side of the dotted line and the signal decomposition module marked on the left side of the dotted line are the same module, specifically including:

[0124] A basic signal synthesis module selects signals from the basic signal library for synthesis to obtain a new synthesized signal;

[0125] The neural network training module creates a neural network to be trained, trains the neural network according to the synthesized new signal and the signal involved in the synthesis, and obtains a trained neural network model.

[0126] The signal decomposition module inputs the synthetic signal to be decomposed into the trained neural network model to perform signal decomposition.

[0127] Example 3

[0128] This embodiment discloses a computer storage medium for storing computer instructions. When the computer instructions are executed by a processor, the steps of a synthetic signal decomposition method described in Embodiment 1 are completed.

[0129] The above are only preferred embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

[0130] Although the above describes the specific implementation methods of the present disclosure in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present disclosure. Technical personnel in the relevant field should understand that on the basis of the technical solution of the present disclosure, various modifications or variations that can be made by those skilled in the art without creative work are still within the scope of protection of the present disclosure.

Claims

1. A method for decomposing a synthetic signal, characterized in that: include: Obtaining the synthetic signal to be decomposed and the basic signal library; Input the synthetic signal to be decomposed into the trained neural network model for signal decomposition, and obtain the meta-signals and number sets belonging to the basic signal library after decomposition; The meta-signals obtained after decomposition are combined and verified with the synthetic signals and attributes before decomposition; The neural network model is obtained by training the basic signal library; The basic signal library is an IEEE1641-BSC basic signal library; The training process of the neural network model is: Create the neural network to be trained; Get the basic signal library; Select a basic signal from a basic signal library for synthesis, and obtain a synthesized new signal; Record the synthesized signals and their properties as well as the basic signals and quantities involved in the synthesis; All new signals, i.e. synthesized signals and quantities, are input into the neural network for pre-decomposition and the pre-decomposition results are verified; When the pre-decomposition results of all new signals are correct, the synthesized signals and attributes are input into the neural network training together with the basic signals and quantities involved in the synthesis to establish the network topology relationship, and the neural network model training is completed; Verify the decomposition result; resynthesize the decomposed element signal, and compare the resynthesized signal with the synthesized signal to be decomposed. If they are the same, the verification is passed; if they are different, the verification fails; if the verification is passed, output the decomposition result; If the verification fails, the neural network model is retrained, and the synthetic signal to be decomposed is decomposed by the retrained neural network model until the decomposition result passes the verification; Basic signal synthesis process, the specific steps are: Input IEEE1641-BSC basic signal library; Construct the attribute object of each signal in the basic signal library, use the class method to describe the attribute, and the unit of the specific attribute is in the description; Combine all described objects into a single set; Randomly select n signal objects from the set, and the selected basic signals are the same or different; Combine the signals of these n signal objects by taking the union of the attributes of all signals; Record the synthesized signal and record it into a synthesized signal set; Record the n signals selected to synthesize the signal and record them in the basic signal set; end.

2. A method for decomposing a synthetic signal as claimed in claim 1, characterized in that: The process of verifying the pre-decomposition results is: The pre-decomposition signal is compared with the basic signal selected for synthesizing the new signal. If they are consistent, the pre-decomposition result is correct. If they are inconsistent, the pre-decomposition result is wrong.

3. A method for decomposing a synthetic signal as claimed in claim 2, characterized in that: When the pre-decomposition result is wrong, the new signal is pre-decomposed again until the pre-decomposition result is verified to be correct.

4. A method for decomposing a synthetic signal as claimed in claim 1, characterized in that: The properties of all selected basic signals are combined for signal synthesis.

5. A system for decomposing a synthetic signal, characterized in that: include: A basic signal synthesis module selects signals from the basic signal library for synthesis to obtain a new synthesized signal; A neural network training module creates a neural network to be trained, trains the neural network according to the synthesized new signal and the signal involved in the synthesis, and obtains a trained neural network model; The signal decomposition module inputs the synthetic signal to be decomposed into the trained neural network model to decompose the signal and obtain the meta-signals and number sets belonging to the basic signal library; The meta-signals obtained after decomposition are combined and verified with the synthetic signals and attributes before decomposition; The basic signal library is an IEEE1641-BSC basic signal library; The training process of the neural network model is: Create the neural network to be trained; Get the basic signal library; Select a basic signal from a basic signal library for synthesis, and obtain a synthesized new signal; Record the synthesized signals and their properties as well as the basic signals and quantities involved in the synthesis; All new signals, i.e. synthesized signals and quantities are input into the neural network for pre-decomposition and the pre-decomposition results are verified; When the pre-decomposition results of all new signals are correct, the synthesized signals and attributes are input into the neural network training together with the basic signals and quantities involved in the synthesis to establish the network topology relationship, and the neural network model training is completed; Verify the decomposition result; resynthesize the decomposed element signal, and compare the resynthesized signal with the synthesized signal to be decomposed. If they are the same, the verification is passed; if they are different, the verification fails; if the verification is passed, output the decomposition result; If the verification fails, the neural network model is retrained, and the synthetic signal to be decomposed is decomposed by the retrained neural network model until the decomposition result passes the verification; Basic signal synthesis process, the specific steps are: Input IEEE1641-BSC basic signal library; Construct the attribute object of each signal in the basic signal library, use the class method to describe the attribute, and the unit of the specific attribute is in the description; Combine all described objects into a single set; Randomly select n signal objects from the set, and the selected basic signals are the same or different; Combine the signals of these n signal objects by taking the union of the attributes of all signals; Record the synthesized signal and record it into a synthesized signal set; Record the n signals selected to synthesize the signal and record them in the basic signal set; end.

6. A computer storage medium, characterized in that: Used to store computer instructions, which, when executed by a processor, complete the steps of a synthetic signal decomposition method as described in any one of claims 1 to 4.

Images