Method, apparatus, device, and storage medium for identifying connection states of single-molecule junctions

Abstract

The present application discloses a method, device, equipment and storage medium for identifying the connection state of a single-molecule junction. The present application relates to the technical field of electrical detection. The method includes: inputting the target on-off experiment data of the target single-molecule junction into a first classification model to obtain the on-off data of the gold-gold junction and the on-off data of the molecule junction in the target on-off experiment data; inputting the target voltammetry experiment data of the target single-molecule junction into a second classification model to obtain the voltammetry data of the gold-gold junction and the voltammetry data of the molecule junction in the target voltammetry experiment data; and identifying the connection state of the target single-molecule junction based on the on-off data of the gold-gold junction, the on-off data of the molecule junction, the voltammetry data of the gold-gold junction and the voltammetry data of the molecule junction. The present application can improve the identification efficiency of the connection state of the single-molecule junction.

Description

Technical Field

[0001] This application relates to the field of electrical detection technologies, and particularly to a method, device, equipment, and storage medium for identifying the connection state of a single-molecule junction. Background Art

[0002] Single-molecule device electronics mainly studies the construction of electronic devices by using a single molecule or a single-molecule film as a conductive channel. Among them, single-molecule electrical detection is the core content of single-molecule device electronics, and usually, the connection state of a single molecule is identified by using the conductance change between the molecule and the electrode.

[0003] Currently, single-molecule electrical detection usually adopts a single-molecule electrical statistical analysis method. By statistically analyzing a large amount of single-molecule electrical measurement data, the statistical laws of single-molecule electrical properties are obtained to infer the connection state of the single-molecule junction. However, traditional statistical methods rely on manual intervention and have the problem of low efficiency in identifying the connection state of single molecules.

[0004] Therefore, how to improve the efficiency of identifying the connection state of a single-molecule junction is an urgent problem to be solved currently. Summary of the Invention

[0005] The main purpose of this application is to provide a method, device, equipment, and storage medium for identifying the connection state of a single-molecule junction, aiming to improve the efficiency of identifying the connection state of a single-molecule junction.

[0006] To achieve the above purpose, this application provides a method for identifying the connection state of a single-molecule junction. The method for identifying the connection state of a single-molecule junction includes:

[0007] Inputting the target on-off experiment data of the target single-molecule junction into a first classification model to obtain the on-off data of the gold-gold junction and the on-off data of the molecule junction in the target on-off experiment data. Among them, the first classification model is trained with the first on-off experiment data of the gold-gold junction as the positive sample and the second on-off experiment data of an unknown structure as the unlabeled sample. The unknown structure is a gold-gold junction or a single-molecule junction. The gold-gold junction refers to a structure formed by directly connecting two gold electrodes, and the single-molecule junction refers to a structure formed by indirectly connecting two gold electrodes through a single molecule;

[0008] Inputting the target volt-ampere experiment data of the target single-molecule junction into a second classification model to obtain the volt-ampere data of the gold-gold junction and the volt-ampere data of the molecule junction in the target volt-ampere experiment data. Among them, the second classification model is trained with the first volt-ampere experiment data of the gold-gold junction as the positive sample and the second volt-ampere experiment data of an unknown structure as the unlabeled sample;

[0009] Based on the gold-gold junction on-off data, the molecular junction on-off data, the gold-gold junction volt-ampere data, and the molecular junction volt-ampere data, identify the connection state of the target single-molecule junction.

[0010] In one embodiment, the method further includes:

[0011] Obtain the first on-off experimental data of the gold-gold junction, and mark the first on-off experimental data as positive samples;

[0012] Obtain the second on-off experimental data of an unknown structure, and determine whether the unknown structure is a single-molecule junction based on the second on-off experimental data;

[0013] When the unknown structure is a single-molecule junction, mark the second on-off experimental data as negative samples;

[0014] Construct a sample data set based on the marked first on-off experimental data and the marked second on-off experimental data, and train a preset model based on the sample data set to obtain the first classification model.

[0015] In one embodiment, the step of determining whether the unknown structure is a single-molecule junction based on the second on-off experimental data includes:

[0016] Perform feature training on a preset autoencoder based on the first on-off experimental data to obtain a target autoencoder;

[0017] Determine the mean square error of the second on-off experimental data through the target autoencoder;

[0018] When the mean square error is greater than a preset threshold, determine that the unknown structure is a single-molecule junction.

[0019] In one embodiment, the method further includes:

[0020] Obtain the first volt-ampere experimental data of the gold-gold junction, and mark the first volt-ampere experimental data as positive samples;

[0021] Obtain the second volt-ampere experimental data of the unknown structure;

[0022] Based on the second volt-ampere experimental data and the marked first volt-ampere experimental data, perform iterative training on a preset initial classifier to obtain multiple weak classifiers;

[0023] Combine the weak classifiers to obtain a strong classifier, and use the strong classifier as the second classification model.

[0024] In one embodiment, the step of combining the weak classifiers to obtain a strong classifier includes:

[0025] Detect the number of the weak classifiers;

[0026] When the number of the weak classifiers reaches a preset number, execute the step of combining the weak classifiers to obtain a strong classifier.

[0027] In one embodiment, the step of identifying the connection state of the target single-molecule junction based on the gold-gold junction on-off data, the molecule-junction on-off data, the gold-gold junction volt-ampere data, and the molecule-junction volt-ampere data includes:

[0028] Classify the molecule-junction on-off data according to the conductance level to obtain on-off data with multiple different conductance levels;

[0029] Classify the molecule-junction volt-ampere data according to the linearity degree to obtain volt-ampere data with multiple different non-linearity degrees;

[0030] Identify the connection state of the target single-molecule junction based on the gold-gold junction on-off data, each on-off data, the gold-gold junction volt-ampere data, and each volt-ampere data.

[0031] In addition, to achieve the above object, the present application further provides a single-molecule junction connection state identification device, and the single-molecule junction connection state identification device includes:

[0032] An on-off data analysis module, configured to input the target on-off experimental data of the target single-molecule junction into a first classification model to obtain the gold-gold junction on-off data and the molecule-junction on-off data in the target on-off experimental data, wherein the first classification model is trained with the first on-off experimental data of the gold-gold junction as a positive sample and the second on-off experimental data of an unknown structure as an unlabeled sample, the unknown structure is a gold-gold junction or a single-molecule junction, the gold-gold junction refers to a structure formed by directly connecting two gold electrodes, and the single-molecule junction refers to a structure formed by indirectly connecting two gold electrodes through a single molecule;

[0033] A volt-ampere data analysis module, configured to input the target volt-ampere experimental data of the target single-molecule junction into a second classification model to obtain the gold-gold junction volt-ampere data and the molecule-junction volt-ampere data in the target volt-ampere experimental data, wherein the second classification model is trained with the first volt-ampere experimental data of the gold-gold junction as a positive sample and the second volt-ampere experimental data of an unknown structure as an unlabeled sample;

[0034] An identification module, configured to identify the connection state of the target single-molecule junction based on the gold-gold junction on-off data, the molecule-junction on-off data, the gold-gold junction volt-ampere data, and the molecule-junction volt-ampere data.

[0035] In addition, to achieve the above object, the present application further provides a storage medium, which is a computer-readable storage medium. A program for implementing the method for identifying the connection state of a single-molecule junction is stored on the computer-readable storage medium. The program for implementing the method for identifying the connection state of a single-molecule junction is executed by a processor to implement the steps of the method for identifying the connection state of a single-molecule junction as described above.

[0036] In addition, to achieve the above object, the present application further provides a computer program product, including a computer program, and when the computer program is executed by a processor, it implements the steps of the method for identifying the connection state of a single-molecule junction as described above.

[0037] The present application provides a method for identifying the connection state of a single-molecule junction. First, the target on-off experiment data of the target single-molecule junction is input into a first classification model to obtain the on-off data of the gold-gold junction and the on-off data of the molecule junction in the target on-off experiment data. Among them, the first classification model is pre-trained with the first on-off experiment data of the gold-gold junction as positive samples and the second on-off experiment data of unknown structures as unlabeled samples. The unknown structure is a gold-gold junction or a single-molecule junction. A gold-gold junction refers to a structure formed by directly connecting two gold electrodes, and a single-molecule junction refers to a structure formed by indirectly connecting two gold electrodes through a single molecule; and, the target voltammetry experiment data of the target single-molecule junction is input into a second classification model to obtain the voltammetry data of the gold-gold junction and the voltammetry data of the molecule junction in the target voltammetry experiment data. Among them, the second classification model is pre-trained with the first voltammetry experiment data of the gold-gold junction as positive samples and the second voltammetry experiment data of unknown structures as unlabeled samples; then, based on the on-off data of the gold-gold junction, the on-off data of the molecule junction, the voltammetry data of the gold-gold junction, and the voltammetry data of the molecule junction, the connection state of the target single-molecule junction is identified.

[0038] In summary, it can be seen that in the present application, the on-off data of the gold-gold junction and the on-off data of the molecule junction in the target on-off experiment data are separated by the pre-trained first classification model, and the voltammetry data of the gold-gold junction and the voltammetry data of the molecule junction in the target voltammetry experiment data are separated by the pre-trained second classification model, so as to analyze and identify the connection state of the single-molecule junction based on the separated data. In this way, compared with the traditional method of analyzing the connection state of a single-molecule junction by manually counting a large number of single-molecule junction measurement data based on statistical laws, the present application realizes data classification by using semi-supervised learning technology, thereby improving the identification efficiency of the connection state of a single-molecule junction. Description of the Drawings

[0039] The drawings here are incorporated into the specification and constitute a part of this specification, showing embodiments consistent with the present application, and are used together with the specification to explain the principles of the present application.

[0040] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the accompanying drawings required in the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0041] Figure 1 It is a schematic flowchart of the first embodiment of the method for identifying the connection state of a single-molecule junction in the present application;

[0042] Figure 2 It is a schematic diagram of on-off experiment data involved in an embodiment of the method for identifying the connection state of a single-molecule junction in the present application;

[0043] Figure 3 It is a schematic diagram of volt-ampere experiment data involved in an embodiment of the method for identifying the connection state of a single-molecule junction in the present application;

[0044] Figure 4 It is a schematic diagram of single-molecule junction data involved in an embodiment of the method for identifying the connection state of a single-molecule junction in the present application;

[0045] Figure 5 It is a schematic flowchart of the process for identifying the connection state of a single-molecule junction involved in an embodiment of the method for identifying the connection state of a single-molecule junction in the present application;

[0046] Figure 6 It is a schematic diagram of the module structure of the device for identifying the connection state of a single-molecule junction in the present application;

[0047] Figure 7 It is a schematic diagram of the device structure of the hardware operating environment involved in the method for identifying the connection state of a single-molecule junction in the embodiments of the present application.

[0048] The realization of the purpose, functional features, and advantages of the present application will be further described with reference to the embodiments and the accompanying drawings. Specific Embodiments

[0049] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0050] To better understand the technical solutions of the present application, the following will be described in detail in combination with the accompanying drawings of the specification and specific embodiments.

[0051] The main solution of this application is: input the target on-off experimental data of the target single-molecule junction into the first classification model to obtain the on-off data of the gold-gold junction and the on-off data of the molecule junction in the target on-off experimental data. Among them, the first classification model is trained with the first on-off experimental data of the gold-gold junction as the positive sample and the second on-off experimental data of unknown structure as the unlabeled sample. The unknown structure is a gold-gold junction or a single-molecule junction. The gold-gold junction refers to a structure formed by directly connecting two gold electrodes, and the single-molecule junction refers to a structure formed by indirectly connecting two gold electrodes through a single molecule; input the target volt-ampere experimental data of the target single-molecule junction into the second classification model to obtain the volt-ampere data of the gold-gold junction and the volt-ampere data of the molecule junction in the target volt-ampere experimental data. Among them, the second classification model is trained with the first volt-ampere experimental data of the gold-gold junction as the positive sample and the second volt-ampere experimental data of unknown structure as the unlabeled sample; based on the on-off data of the gold-gold junction, the on-off data of the molecule junction, the volt-ampere data of the gold-gold junction, and the volt-ampere data of the molecule junction, identify the connection state of the target single-molecule junction.

[0052] Currently, single-molecule electrical detection usually adopts the single-molecule electrical statistical analysis method. By statistically analyzing a large amount of single-molecule electrical measurement data, the statistical laws of single-molecule electrical properties are obtained to infer the connection state of the single-molecule junction. However, the traditional statistical method relies on manual intervention and has the problem of low efficiency in identifying the single-molecule connection state.

[0053] Therefore, how to improve the efficiency of identifying the connection state of the single-molecule junction is an urgent problem to be solved currently.

[0054] This application separates the on-off data of the gold-gold junction and the on-off data of the molecule junction in the target on-off experimental data through the pre-trained first classification model, and separates the volt-ampere data of the gold-gold junction and the volt-ampere data of the molecule junction in the target volt-ampere experimental data through the pre-trained second classification model, so as to analyze and identify the connection state of the single-molecule junction based on the separated data. In this way, compared with the traditional method of statistically analyzing a large amount of single-molecule junction measurement data based on statistical laws to analyze the connection state of the single-molecule junction, this application realizes data classification by using semi-supervised learning technology, thereby improving the efficiency of identifying the connection state of the single-molecule junction.

[0055] It should be noted that the execution subject of the method in each embodiment of the single-molecule junction connection state recognition method of this application can be an identification system, or a computing service device with data processing, network communication, and program running functions, such as a tablet computer, a personal computer, a mobile phone, etc., or a single-molecule junction connection state recognition device capable of realizing the above functions. This embodiment does not make specific limitations on this. Hereinafter, taking the identification system as the execution subject as an example, this embodiment and the following embodiments will be described.

[0056] Based on this, the present application proposes a method for identifying the connection state of a single-molecule junction in the first embodiment. Please refer to Figure 1 , the method for identifying the connection state of the single-molecule junction includes steps S10 to S30:

[0057] Step S10, input the target on-off experiment data of the target single-molecule junction into the first classification model to obtain the on-off data of the gold-gold junction and the on-off data of the molecule junction in the target on-off experiment data. Among them, the first classification model is trained with the first on-off experiment data of the gold-gold junction as the positive sample and the second on-off experiment data of the unknown structure as the unlabeled sample. The gold-gold junction refers to a structure formed by directly connecting two gold electrodes, and the single-molecule junction refers to a structure formed by indirectly connecting two gold electrodes through a single molecule;

[0058] It should be noted that the on-off experiment data refers to the I-Z data obtained when performing an on-off experiment on a single-molecule junction or a gold-gold junction, that is, the current-displacement data, which can be understood as the data of the current between two electrodes changing with the distance between the two electrodes.

[0059] In this embodiment, the gold-gold junction refers to a structure formed by connecting a gold tip and a gold substrate without grown molecules. Among them, the gold tip and the gold substrate can be understood as two gold electrodes. The gold tip is the electrode in the shape of a tip among the two gold electrodes, and the gold substrate without grown molecules refers to a gold-based substrate material without attached molecules on its surface. It can be understood that the gold-gold junction refers to a structure formed by directly connecting two gold electrodes. When the two electrodes approach each other and are in direct contact without an intermediate molecule as a connecting bridge, a gold-gold junction is formed. The single-molecule junction refers to a structure formed by connecting a gold tip and a gold substrate with grown molecules. The gold substrate with grown molecules refers to a gold-based substrate material with a single molecule attached to its surface. It can be understood that the single-molecule junction refers to a structure formed by indirectly connecting the gold tip through the single molecule on the gold substrate to the gold substrate.

[0060] The single-molecule junction to be identified is called the target single-molecule junction for distinction. Obtain the on-off experiment data of the target single-molecule junction (hereinafter referred to as the target on-off experiment data for distinction) and input it into a pre-trained classification model (hereinafter referred to as the first classification model for distinction) to obtain the on-off data of the gold-gold junction and the on-off data of the molecule junction in the target on-off experiment data, that is, separate the on-off data of the gold-gold junction and the on-off data of the molecule junction in the target on-off experiment data. Among them, the first classification model is trained with the on-off experiment data of the gold-gold junction (hereinafter referred to as the first on-off experiment data for distinction) as the positive sample and the on-off experiment data of the unknown structure (hereinafter referred to as the second on-off experiment data for distinction) as the unlabeled sample. The unknown structure can be a gold-gold junction or a single-molecule junction.

[0061] In a feasible implementation manner, first, a small amount of first on-off experiment data and a large amount of second on-off experiment data are obtained, and data preprocessing such as data cleaning, denoising, conductivity conversion, and normalization is performed on these data to ensure the quality and consistency of the data. Then, a feature selection method based on statistical histogram distribution or physical characteristics is used to convert the data to the same data dimension, and according to actual requirements, machine learning techniques such as principal component analysis and non-negative matrix factorization are used to further extract features or reduce the dimensional space of the processed I-Z data; then, based on the first classification model, the gold-gold junction on-off data and the molecular junction on-off data in the unknown single-molecule junction connection state on-off experiment data (i.e., the target on-off experiment data) are distinguished. Specifically, methods such as PUL (Positive and Unlabeled Learning)-naive standard classification method and PUL-Two Step Strategy can be used according to actual requirements.

[0062] Step S20: Input the target volt-ampere experiment data of the target single-molecule junction into the second classification model to obtain the gold-gold junction volt-ampere data and the molecular junction volt-ampere data in the target volt-ampere experiment data, where the second classification model is trained with the first volt-ampere experiment data of the gold-gold junction as positive samples and the second volt-ampere experiment data of unknown structures as unlabeled samples.

[0063] It should be noted that the volt-ampere experiment data refers to the I-V data obtained during the volt-ampere experiment on a single-molecule junction or a gold-gold junction, that is, the current-voltage data, which can be understood as the data where the current between two electrodes changes with the voltage between the two electrodes.

[0064] Obtain the volt-ampere experiment data of the target single-molecule junction (hereinafter referred to as the target volt-ampere experiment data for distinction), and input the target volt-ampere experiment data into a pre-trained classification model (hereinafter referred to as the second classification model for distinction) to obtain the gold-gold junction volt-ampere data and the molecular junction volt-ampere data in the target volt-ampere experiment data. Among them, the second classification model is pre-trained with the volt-ampere experiment data of the gold-gold junction (hereinafter referred to as the first volt-ampere experiment data for distinction) as positive samples and the volt-ampere experiment data of unknown structures (hereinafter referred to as the second volt-ampere experiment data for distinction) as unlabeled samples.

[0065] In a feasible embodiment, a small amount of first voltammetric experimental data and a large amount of second voltammetric experimental data are obtained. It should be understood that the first voltammetric experimental data has a linear curve feature, and the curve feature of the second voltammetric experimental data is non-linear, and the degree of non-linearity is different; then, the obtained I-V data is preprocessed such as data cleaning, denoising, conductance conversion, and normalization; the second classification model obtained based on semi-supervised learning separates the voltammetric data of the gold-gold junction and the voltammetric data of the molecular junction in the I-V data (i.e., the target voltammetric experimental data) of the unknown single-molecule junction connection state. Specifically, methods such as the PUL-naive standard classification method and PUL-PU Bagging (PUL-Positive-Unlabeled Bagging) can be used according to actual needs.

[0066] Thus, in the embodiment of the present application, by using unsupervised / semi-supervised machine learning methods, based on the electrical data of a single gold-gold junction sample, the I-Z (current-displacement) data and I-V (current-voltage) data of the on-off experiment in single-molecule electrical detection are analyzed and identified, so that the different I-Z data and I-V data form different corresponding relationships, and single-molecule junctions in different connection states (gold-gold junctions (gold-gold), molecular junctions with different conductance levels (gold-molecule-gold)) are distinguished, providing reliable data support for the subsequent research of single-molecule device electronics.

[0067] Step S30: Based on the gold-gold junction on-off data, the molecular junction on-off data, the gold-gold junction voltammetric data, and the molecular junction voltammetric data, identify the connection state of the target single-molecule junction.

[0068] It should be noted that the specific significance of identifying the single-molecule junction connection state is reflected in the following aspects: understanding the electron transport mechanism, by identifying the connection state of the single-molecule junction, the electron transport mechanism inside and between molecules can be deeply studied, and the relationship between the molecular structure and electron transport characteristics can be revealed; optimizing the performance of nano-devices, identifying the connection state of the single-molecule junction can help design and optimize nano-scale electronic devices, such as molecular wires, molecular switches, etc., to improve their performance and stability. By precisely controlling the connection state of the molecular junction, highly sensitive and highly selective molecular sensors can be developed for detecting trace substances; new material development, identifying the connection state of the single-molecule junction helps to design and synthesize new functional materials with specific electronic properties, such as conductive polymers, self-assembled monolayers, etc. Understanding the interface connection state between the molecule and the electrode can optimize the interface characteristics of the material and improve the overall performance of the material. Thus, identifying the single-molecule junction connection state has important specific significance in aspects such as electron transport mechanism research, nano-device optimization, and new material development, and can promote the development of multiple scientific and technological fields.

[0069] Based on the on-off data of the metal-metal junction, the on-off data of the molecular junction, the volt-ampere data of the metal-metal junction, and the volt-ampere data of the molecular junction, the connection state of the target single-molecule junction is analyzed and obtained.

[0070] In a feasible implementation manner, after obtaining various types of data output by the classification model, the on-off data of the molecular junction output by the first classification model is clustered into categories with different high and low conductivities. Specifically, clustering methods such as KMeans (K-Means Clustering) and spectral clustering can be selected according to actual needs; and the volt-ampere data of the molecular junction output by the second classification model is clustered into categories with different degrees of non-linearity. Specifically, clustering methods such as spectral clustering and HDBSCAN (Hierarchical Density-Based Spatial Clustering of Applications with Noise) can be selected according to actual needs. The I-Z data curves with different conductivity levels and the I-V data curves with different degrees of non-linearity are made to correspond one by one. Based on the correspondence between I-Z and I-V at high and low conductivities of the molecular junction, and according to the theoretical basis of molecular chemistry, the interpretability of single-molecule junctions with different connection states is improved from aspects such as the connection mode between the molecule and the electrode (covalent bond connection, non-covalent bond connection) and the conformational change of the molecule (extended conformation, bent or folded conformation), and the relationship between conductivity and the connection state of the single-molecule junction is further understood.

[0071] The embodiments of this application utilize the characteristics of unsupervised learning. Without the need for a large amount of labeled data, by mining the statistical laws and similarities within the conductivity data, potential connection patterns and features can be discovered, realizing the effective identification of the connection state of the single-molecule junction. Compared with traditional statistical methods, this unsupervised learning method reduces the need for human intervention, reduces the influence of human errors and subjectivity on the identification results, has a faster calculation speed and higher identification efficiency, which is particularly important for real-time or online applications of identifying the connection state of single-molecule junctions. Moreover, through the analysis of unsupervised learning algorithms, the distribution and characteristic differences of different connection states of single-molecule junctions can be more intuitively understood, which helps researchers obtain valuable physical and chemical information from the data and provides guidance for further theoretical research and experimental design.

[0072] Specifically, the single-molecule junction connection state identification method provided by the embodiment of the present application has the advantages of strong adaptability, low data annotation dependence, ability to discover new connection states and strong real-time processing capabilities. It is understandable that the unsupervised learning method does not rely on a specific data distribution hypothesis, can automatically adapt to data changes, can effectively identify the single-molecule junction connection state under different experimental environments and data characteristics, and has stronger versatility and robustness; in the study of single-molecule junctions, due to the complexity of the microscopic state, the annotation work is difficult and error-prone, and the unsupervised learning method directly analyzes the original data, reduces labor costs and annotation errors, and improves data utilization efficiency; the unsupervised learning method, relying on its ability to mine the intrinsic structure of the data, can discover potential, unrecognized single-molecule junction connection state patterns from the data, open up new directions for the study of single-molecule junctions, and discover new physical phenomena and laws; in some application scenarios of real-time monitoring of the connection state of single-molecule junctions, unsupervised learning can quickly process newly collected data and output the judgment results of the connection state in a timely manner. This real-time nature makes the unsupervised learning method have obvious advantages in dynamically monitoring the behavior of single-molecule junctions and real-time feedback of experimental results.

[0073] Based on the first embodiment of the present application, in the second embodiment of the present application, the same or similar contents as those in the above-mentioned embodiment 1 can be referred to the above introduction, and will not be repeated in the following. On this basis, the single molecule junction connection state identification method of the present application also includes:

[0074] Step A10, obtaining first on-off experimental data of a gold-gold junction, and marking the first on-off experimental data as a positive sample;

[0075] It should be noted that if Figure 2 The figure shows the schematic diagram of the on-off experimental data, where: Figure 2 (a) is the on-off experimental data A1 (i.e., the first on-off experimental data) of the gold needle tip and the gold substrate without molecular growth. It can be understood that since there are no molecules involved in the circuit connection, the data A1 is relatively pure (the on-off curve is smooth and does not produce conductivity steps).

[0076] The first on-off experimental data of the gold-gold junction is obtained, and the first on-off experimental data is marked as a positive sample for subsequent classification model to perform semi-supervised learning.

[0077] Step A20, obtaining second on-off experimental data of the unknown structure, and determining whether the unknown structure is a single-molecule junction based on the second on-off experimental data;

[0078] It should be noted that Figure 2In Figure (b), A2 (i.e., the second on-off experiment data) is the on-off experiment data of the gold needle tip and the grown molecular gold substrate. It can be understood that the data A2 is mixed with various gold-gold junction data and molecular junction data in different connection states, and the data proportions are different, making it difficult for humans to effectively distinguish. After obtaining the first on-off experiment data and the second on-off experiment data, the conductance histogram statistical distribution processing is performed on the data to make the characteristic dimensions of the data consistent, facilitating subsequent model training and learning.

[0079] Obtain the second on-off experiment data of an unknown structure, and this second on-off experiment data belongs to an unlabeled sample. Regarding the unknown structure, it should be understood that the unknown structure is a gold-gold junction or a single-molecule junction. Therefore, based on the second on-off experiment data of the unknown structure, it is determined whether the unknown structure is a single-molecule junction to identify the single-molecule junction in the unknown structure.

[0080] In this embodiment, the step A20 may include:

[0081] Step A201, perform feature training on a preset autoencoder based on the first on-off experiment data to obtain a target autoencoder;

[0082] Step A202, determine the mean squared error of the second on-off experiment data through the target autoencoder;

[0083] Step A203, when the mean squared error is greater than a preset threshold, determine that the unknown structure is a single-molecule junction.

[0084] It should be noted that the MSE (Mean Squared Error) threshold for screening suspected negative samples is set in advance according to specific problems and data characteristics, that is, the above-mentioned preset threshold.

[0085] The steps for determining whether an unknown structure is a single-molecule junction include: first, perform feature training on a preset autoencoder based on the first on-off experiment data to obtain a trained autoencoder (hereinafter referred to as the target autoencoder for distinction); then determine the mean squared error of the second on-off experiment data through the target autoencoder; when the mean squared error is greater than a preset threshold, determine that the unknown structure is a single-molecule junction.

[0086] Exemplarily, when processing I-Z data, the PUL-Two Step Strategy method is used to perform data learning in two steps. Specifically, the first step is to identify suspected negative samples: the I-Z data A1 and the I-Z data A2 are respectively identified as positive samples and unlabeled samples, and the initial model is determined as an autoencoder; the autoencoder is used to perform feature training on the positive samples, and the MSE error is used as the optimization object to obtain a trained autoencoder (i.e., the target autoencoder); the MSE error of each unlabeled sample in the unlabeled sample set is calculated through the target autoencoder, and the MSE error corresponding to each unlabeled sample is obtained; the unlabeled samples with an MSE error greater than the preset threshold t are labeled as suspected negative samples R.

[0087] Step A30, in the case where the unknown structure is a single-molecule junction, label the second on-off experiment data as a negative sample;

[0088] In the case where the unknown structure is a single-molecule junction, label the second on-off experiment data of the unknown structure as a negative sample. That is, identify suspected negative samples from the unlabeled samples.

[0089] Step A40, construct a sample data set based on the labeled first on-off experiment data and the labeled second on-off experiment data, and train a preset model based on the sample data set to obtain the first classification model.

[0090] It should be noted that the preset model can be an LGBM (Light GradientBoosting Machine) classifier based on gradient boosting decision trees.

[0091] Construct a sample data set based on the first on-off experiment data labeled with positive sample labels and the second on-off experiment data labeled with negative sample labels, so as to train a preset model based on the sample data set to obtain the first classification model.

[0092] Exemplarily, the second step is to train the final classifier: combine the positive sample set and the suspected negative sample set R to obtain a new sample data set D. Among them, in the sample data set D, positive class labels 1 are assigned to positive samples, and negative class labels -1 are assigned to suspected negative samples. That is to say, the sample data set D contains a labeled data set, which can be used for training traditional supervised learning algorithms; then, according to the problem characteristics and data properties, select the LGBM classifier based on gradient boosting decision trees as the final classification model (i.e., the preset model); train the preset model based on the sample data set D. Since there are differences in the quantity between the positive sample set and the suspected negative sample set R, sample imbalance weights are introduced during the training process of the preset model to improve the accuracy of the model's classification prediction results; after obtaining the trained classification model (i.e., the first classification model), this first classification model can be used to predict unlabeled samples to obtain gold-gold junction data and molecule-junction data.

[0093] Specifically, input the data A2 shown in (b) in Figure 2 into the first classification model to obtain Figure 2 the gold-gold junction on-off data A3 shown in (c) in Figure 2 and the molecule-junction on-off data A4 shown in (d) in

[0094] In this way, the embodiment of the present application identifies negative samples in the unlabeled sample set by using an autoencoder, constructs a sample data set based on the negative samples and positive samples, and provides the sample data set for the preset model to perform semi-supervised learning to obtain the first classification model, so as to separate the gold-gold junction on-off data and the molecule-junction on-off data through the first classification model.

[0095] In this embodiment, the single molecule-junction connection state recognition method of the present application may further include:

[0096] Step B10, obtain the first voltammetric experimental data of the gold-gold junction, and label the first voltammetric experimental data as positive samples;

[0097] It should be noted that, as Figure 3 shown in the schematic diagram of voltammetric experimental data, where Figure 3 the voltammetric experimental data B1 (i.e., the first voltammetric experimental data) of the gold needle tip and the ungrown molecular gold substrate in (a) in

[0098] can be understood. Since only the gold-gold junction participates in the circuit connection, the data B1 is relatively linear.

[0099] Obtain the first voltammetric experimental data of the gold-gold junction, and label the first voltammetric experimental data as positive samples for subsequent semi-supervised learning of the classification model.

[0100] It should be noted thatFigure 3 As shown in (b) in the text, the voltammetry experimental data B2 (i.e., the second voltammetry experimental data) of the gold needle tip and the grown molecular gold substrate are presented. It can be understood that since the data B2 is mixed with various Au-Au junction I-V data and molecular junction I-V data at different conductance levels, and the proportions are different, it is difficult for humans to effectively distinguish them.

[0101] Obtain the second voltammetry experimental data of the unknown structure for subsequent model training.

[0102] Step B30: Based on the second voltammetry experimental data and the labeled first voltammetry experimental data, perform iterative training on a preset initial classifier to obtain multiple weak classifiers;

[0103] Based on the second voltammetry experimental data and the first voltammetry experimental data labeled with positive sample labels, perform iterative training on a preset initial classifier to obtain the weak classifiers generated after each training.

[0104] Step B40: Combine the weak classifiers to obtain a strong classifier, and use the strong classifier as the second classification model.

[0105] Combine the weak classifiers to obtain a strong classifier, and use this strong classifier as the second classification model.

[0106] In this embodiment, the step B40 may include:

[0107] Step B401: Detect the number of the weak classifiers;

[0108] Step B402: When the number of the weak classifiers reaches a preset number, perform the step of combining the weak classifiers to obtain a strong classifier.

[0109] It should be noted that the number of weak classifiers in the ensemble learning is preset to be a preset number T in advance.

[0110] Monitor the iterative training process of the classifier, judge whether the number of weak classifiers reaches the preset number, and when the number of weak classifiers reaches the preset number, combine the weak classifiers to obtain a strong classifier.

[0111] Exemplarily, for the I-V data, the PUL-PU Bagging method is used for data learning: First, Figure 3 the Au-Au junction data B1 (i.e., the first voltammetry experimental data) shown in (a) in the text and Figure 3The I-V data B2 shown in (b) (i.e., the second volt-ampere experimental data) are respectively identified as positive samples P and unlabeled samples U, where the number of positive samples is Np and the number of unlabeled samples is Nu. At the same time, determine the number T of weak classifiers in the ensemble learning, and the basic classification algorithm is the decision tree; then, based on P and U, iteratively train the weak classifiers. Specifically, for each iteration t = 1, 2,..., T, perform sampling with replacement on P and U to obtain Pt and Ut, and label Pt as the positive class 1 and Ut as the negative class -1. Use the selected decision tree algorithm to train the training subset Dt = Pt ∪ Ut to obtain a weak classifier Ht; after T iterations of training, obtain T weak classifiers, which are respectively denoted as H1, H2,..., HT, and then combine these weak classifiers into a strong classifier H through the majority voting method. This strong classifier is the second classification model.

[0112] Specifically, input Figure 3 the data B2 shown in (b) into the second classification model to obtain Figure 3 the volt-ampere data B3 of the metal-metal junction shown in (c) and Figure 3 the volt-ampere data B4 of the molecular junction shown in (d).

[0113] In this way, through the embodiment of the present application, by iteratively training weak classifiers based on positive samples and unlabeled samples, combining the weak classifiers obtained after multiple rounds of iterative training to obtain a strong classifier, and using this strong classifier as the second classification model, the separation of the volt-ampere data of the metal-metal junction and the volt-ampere data of the molecular junction is realized.

[0114] In this embodiment, the step S30 may include:

[0115] Step S301, classify the on-off data of the molecular junction according to the conductance level to obtain on-off data with multiple different conductance levels;

[0116] After obtaining the on-off data of the molecular junction, classify the on-off data of the molecular junction according to the conductance level to obtain on-off data with multiple different conductance levels.

[0117] Exemplarily, use the spectral clustering method to classify the on-off data A4 of the molecular junction into high and low conductance levels to obtain Figure 2 the low-conductance on-off data A5 shown in (e) and Figure 2 the high-conductance on-off data A6 shown in (f).

[0118] Step S302, classify the volt-ampere data of the molecular junction according to the linearity degree to obtain volt-ampere data with multiple different non-linearity degrees;

[0119] After obtaining the volt-ampere data of the molecular junction, classify the volt-ampere data of the molecular junction according to the linearity degree to obtain volt-ampere data with multiple different non-linearity degrees.

[0120] Exemplarily, the HDBSCAN method is used to classify the linearity of the I-V data B4 of different molecular junctions with different linearity degrees, and the Figure 3 strongly non-linear I-V data B5 shown in (e) in Figure 3 and the weakly non-linear I-V data B6 shown in (f) in

[0121] Step S303: Based on the gold-gold junction on-off data, each of the on-off data, the gold-gold junction volt-ampere data, and each of the volt-ampere data, identify the connection state of the target single-molecule junction.

[0122] Based on the gold-gold junction on-off data, the on-off data of different conductance levels, the gold-gold junction volt-ampere data, and the volt-ampere data of different non-linearity degrees, determine the connection state of the target single-molecule junction.

[0123] Exemplarily, the previously obtained gold-gold junction I-Z data A3 (i.e., the gold-gold junction on-off data), the low-conductance step I-Z data A5, the high-conductance step I-Z data A6, and the gold-gold junction I-V data B3 (i.e., the gold-gold junction volt-ampere data), the strongly non-linear I-V data B5, and the weakly non-linear I-V data B6 are corresponded one by one, that is, as Figure 4 shown in the schematic diagram of single-molecule junction data, A3-B3, A5-B5, A6-B6. Subsequently, based on the theoretical basis of molecular chemistry, the relationship between conductance and the connection state of the single-molecule junction can be further understood from aspects such as the connection mode between the molecule and the electrode (covalent bond connection, non-covalent bond connection) and the conformational change of the molecule (extended conformation, bent or folded conformation).

[0124] In a feasible implementation manner, as Figure 5 shown in the schematic diagram of the single-molecule junction connection state recognition process, the single-molecule connection state recognition includes two parts, namely the I-Z data analysis based on the unsupervised algorithm and the I-V data analysis based on the unsupervised algorithm, aiming to separate the I-Z data into the gold-gold junction I-Z data and the molecular junction I-Z data of different conductance levels, and separate the I-V data into the gold-gold junction I-V data and the molecular junction I-V data of different non-linearity degrees, and perform differential correspondence on the separated I-Z data and the separated I-V data, so as to obtain the connection state of the current single-molecule junction based on the final correspondence relationship and data analysis.

[0125] Thus, based on the unsupervised learning method, the embodiments of the present application provide a good feature representation and clustering basis for data driven by a large number of unlabeled samples, realizing the recognition of the connection state of single-molecule junctions, greatly improving the detection accuracy and efficiency; moreover, unsupervised learning can adapt to datasets of different scales and complexities, and has better adaptability to complex and variable problems such as the recognition of the connection state of single-molecule junctions; and, based on the fast classification ability of unsupervised learning, it is beneficial to build an automated detection platform to accelerate the screening and characterization of new molecular devices.

[0126] The embodiments of the present application further provide a device for recognizing the connection state of a single-molecule junction. Please refer to Figure 6 , the device for recognizing the connection state of a single-molecule junction includes:

[0127] The on-off data analysis module 10 is configured to input the target on-off experimental data of the target single-molecule junction into the first classification model to obtain the on-off data of the gold-gold junction and the on-off data of the molecular junction in the target on-off experimental data. Among them, the first classification model is trained with the first on-off experimental data of the gold-gold junction as the positive sample and the second on-off experimental data of the unknown structure as the unlabeled sample. The unknown structure is a gold-gold junction or a single-molecule junction. The gold-gold junction refers to a structure formed by directly connecting two gold electrodes, and the single-molecule junction refers to a structure formed by indirectly connecting two gold electrodes through a single molecule;

[0128] The voltammetry data analysis module 20 is configured to input the target voltammetry experimental data of the target single-molecule junction into the second classification model to obtain the voltammetry data of the gold-gold junction and the voltammetry data of the molecular junction in the target voltammetry experimental data. Among them, the second classification model is trained with the first voltammetry experimental data of the gold-gold junction as the positive sample and the second voltammetry experimental data of the unknown structure as the unlabeled sample;

[0129] The recognition module 30 is configured to recognize the connection state of the target single-molecule junction based on the on-off data of the gold-gold junction, the on-off data of the molecular junction, the voltammetry data of the gold-gold junction, and the voltammetry data of the molecular junction.

[0130] Optionally, the on-off data analysis module 10 is further configured to:

[0131] Obtain the first on-off experimental data of the gold-gold junction and label the first on-off experimental data as the positive sample;

[0132] Obtain the second on-off experimental data of the unknown structure and determine whether the unknown structure is a single-molecule junction based on the second on-off experimental data;

[0133] In the case that the unknown structure is a single-molecule junction, label the second on-off experimental data as the negative sample;

[0134] Construct a sample data set based on the first on-off experiment data after marking and the second on-off experiment data after marking, and train a preset model based on the sample data set to obtain the first classification model.

[0135] Optionally, the on-off data analysis module 10 is further configured to:

[0136] Perform feature training on a preset autoencoder based on the first on-off experiment data to obtain a target autoencoder;

[0137] Determine the mean square error of the second on-off experiment data through the target autoencoder;

[0138] When the mean square error is greater than a preset threshold, determine that the unknown structure is a single-molecule junction.

[0139] Optionally, the voltammetry data analysis module 20 is further configured to:

[0140] Obtain the first voltammetry experiment data of the Au-Au junction and mark the first voltammetry experiment data as a positive sample;

[0141] Obtain the second voltammetry experiment data of the unknown structure;

[0142] Based on the second voltammetry experiment data and the first voltammetry experiment data after marking, perform iterative training on a preset initial classifier to obtain multiple weak classifiers;

[0143] Combine each of the weak classifiers to obtain a strong classifier, and use the strong classifier as the second classification model.

[0144] Optionally, the voltammetry data analysis module 20 is further configured to:

[0145] Detect the number of the weak classifiers;

[0146] When the number of the weak classifiers reaches a preset number, execute the step of combining each of the weak classifiers to obtain a strong classifier.

[0147] Optionally, the recognition module 30 is further configured to:

[0148] Classify the on-off data of the molecular junction to obtain on-off data at multiple different conductance levels;

[0149] Classify the voltammetry data of the molecular junction according to the degree of linearity to obtain voltammetry data at multiple different non-linearity degrees;

[0150] Based on the on-off data of the Au-Au junction, each of the on-off data, the voltammetry data of the Au-Au junction, and each of the voltammetry data, identify the connection state of the target single-molecule junction.

[0151] The single-molecule junction connection state recognition device provided by the embodiment of the present application adopts the single-molecule junction connection state recognition method in the above embodiment, and can solve the technical problem of how to improve the recognition efficiency of the single-molecule junction connection state. Compared with the prior art, the beneficial effects of the single-molecule junction connection state recognition device provided by the embodiment of the present application are the same as those of the single-molecule junction connection state recognition method provided by the above embodiment, and other technical features in the single-molecule junction connection state recognition device are the same as the features disclosed in the above embodiment method, which will not be elaborated here.

[0152] The present application provides a single-molecule junction connection state recognition device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the single-molecule junction connection state recognition method in the first embodiment above.

[0153] Next, refer to Figure 7 , which shows a schematic structural diagram suitable for implementing the single-molecule junction connection state recognition device of the embodiment of the present application. The single-molecule junction connection state recognition device in the embodiment of the present application may include, but is not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions: tablet computers), PMPs (Portable Media Players), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 7 The single-molecule junction connection state recognition device shown is only an example and should not impose any limitations on the functions and usage scope of the embodiment of the present application.

[0154] As Figure 7As shown, the single-molecule junction connection state recognition device may include a processing device 1001 (such as a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to the program stored in the read-only memory 1002 or the program loaded from the storage device 1003 into the random access memory 1004. In the random access memory 1004, various programs and data required for the operation of the single-molecule junction connection state recognition device are also stored. The processing device 1001, the read-only memory 1002, and the random access memory 1004 are connected to each other through a bus 1005. The input / output interface 1006 is also connected to the bus. Generally, the following systems may be connected to the input / output interface 1006: an input device 1007 including, for example, a touch screen, a touchpad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD: Liquid Crystal Display), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 can allow the single-molecule junction connection state recognition device to communicate with other devices wirelessly or wiredly to exchange data. Although the figure shows a single-molecule junction connection state recognition device with various systems, it should be understood that it is not required to implement or have all the shown systems. More or fewer systems may be implemented or had alternatively.

[0155] In particular, according to the embodiments disclosed in the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains program codes for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication device, or installed from the storage device 1003, or installed from the read-only memory 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiments disclosed in the present application are executed.

[0156] The single-molecule junction connection state recognition device provided by the present application adopts the single-molecule junction connection state recognition method in the above-mentioned embodiment, and can solve the technical problem of how to improve the recognition efficiency of the single-molecule junction connection state. Compared with the prior art, the beneficial effects of the single-molecule junction connection state recognition device provided by the present application are the same as those of the single-molecule junction connection state recognition method provided by the above-mentioned embodiment, and other technical features in the single-molecule junction connection state recognition device are the same as those disclosed in the method of the previous embodiment, and will not be elaborated here.

[0157] It should be understood that each part disclosed in this application can be implemented by hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in a suitable manner in any one or more embodiments or examples.

[0158] As described above, the above are only specific embodiments of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art within the technical scope disclosed in this application can easily think of changes or substitutions, which should all be covered within the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claims.

[0159] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., computer programs) stored thereon, and the computer-readable program instructions are used to execute the single-molecule junction connection state recognition method in the above embodiments.

[0160] The computer-readable storage medium provided by this application can be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems or devices, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections with one or more wires, portable computer disks, hard disks, random access memory (RAM: Random Access Memory), read-only memory (ROM: Read Only Memory), erasable programmable read-only memory (EPROM: Erasable Programmable Read Only Memory or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM: CD-Read Only Memory), optical storage devices, magnetic storage devices, or any suitable combination of the above. In this embodiment, the computer-readable storage medium can be any tangible medium that contains or stores a program, and this program can be used by or in combination with an instruction execution system or device. The program code contained on the computer-readable storage medium can be transmitted by any appropriate medium, including but not limited to: wires, optical cables, RF (RadioFrequency: radio frequency), etc., or any suitable combination of the above.

[0161] The above computer-readable storage medium can be included in the single-molecule junction connection state recognition device; or it can exist separately without being assembled into the single-molecule junction connection state recognition device.

[0162] The above computer-readable storage medium carries one or more programs. When the one or more programs are executed by the single-molecule junction connection state recognition device, the single-molecule junction connection state recognition device is caused to: input the target on-off experiment data of the target single-molecule junction into the first classification model to obtain the on-off data of the gold-gold junction and the on-off data of the molecular junction in the target on-off experiment data, wherein the first classification model is trained with the first on-off experiment data of the gold-gold junction as the positive sample and the second on-off experiment data of the unknown structure as the unlabeled sample, the unknown structure being a gold-gold junction or a single-molecule junction, the gold-gold junction referring to a structure formed by directly connecting two gold electrodes, and the single-molecule junction referring to a structure formed by indirectly connecting two gold electrodes through a single molecule; input the target voltammetry experiment data of the target single-molecule junction into the second classification model to obtain the voltammetry data of the gold-gold junction and the voltammetry data of the molecular junction in the target voltammetry experiment data, wherein the second classification model is trained with the first voltammetry experiment data of the gold-gold junction as the positive sample and the second voltammetry experiment data of the unknown structure as the unlabeled sample; and identify the connection state of the target single-molecule junction based on the on-off data of the gold-gold junction, the on-off data of the molecular junction, the voltammetry data of the gold-gold junction, and the voltammetry data of the molecular junction.

[0163] Computer program code for performing the operations of the present application may be written in one or more programming languages or combinations thereof. The programming languages include object-oriented programming languages such as Java, Smalltalk, C++, and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, executed as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., by using an Internet service provider to connect through the Internet).

[0164] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagram may represent a module, a segment of a program, or a part of code that contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than that marked in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, as well as combinations of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system that performs the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0165] The modules involved in the embodiments described in the present application can be implemented in software or in hardware. Among them, the name of the module does not constitute a limitation on the unit itself in some cases.

[0166] The readable storage medium provided by the present application is a computer-readable storage medium, and the computer-readable storage medium stores computer-readable program instructions (i.e., computer programs) for executing the above-mentioned single-molecule junction connection state recognition method, which can solve the technical problem of how to improve the recognition efficiency of the single-molecule junction connection state. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided by the present application are the same as those of the single-molecule junction connection state recognition method provided by the above embodiments, and will not be elaborated here.

[0167] An embodiment of the present application provides a computer program product, including a computer program, and when the computer program is executed by a processor, it implements the steps of the single-molecule junction connection state recognition method as described above.

[0168] The computer program product provided by the present application can improve the recognition efficiency of the single-molecule junction connection state. Compared with the prior art, the beneficial effects of the computer program product provided by the embodiments of the present application are the same as those of the single-molecule junction connection state recognition method provided by the above embodiments, and will not be elaborated here.

[0169] The above are only the preferred embodiments of the present application, and do not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made by using the specification and drawings of the present application, or directly or indirectly applied in other related technical fields, shall be included in the patent scope of the present application in the same way.

Claims

1. A method for identifying the connection state of a single-molecule junction, characterized in that, The method for identifying the connection state of a single-molecule junction includes: Inputting the target on-off experiment data of the target single-molecule junction into a first classification model to obtain the on-off data of the gold-gold junction and the on-off data of the single-molecule junction in the target on-off experiment data. Among them, the first classification model is trained with the first on-off experiment data of the gold-gold junction as positive samples and the second on-off experiment data of unknown structures as unlabeled samples. The unknown structures are gold-gold junctions or single-molecule junctions. The gold-gold junction refers to a structure formed by directly connecting two gold electrodes, and the single-molecule junction refers to a structure formed by indirectly connecting two gold electrodes through a single molecule; Inputting the target voltammetry experiment data of the target single-molecule junction into a second classification model to obtain the voltammetry data of the gold-gold junction and the voltammetry data of the single-molecule junction in the target voltammetry experiment data. Among them, the second classification model is trained with the first voltammetry experiment data of the gold-gold junction as positive samples and the second voltammetry experiment data of unknown structures as unlabeled samples; Based on the on-off data of the gold-gold junction, the on-off data of the single-molecule junction, the voltammetry data of the gold-gold junction, and the voltammetry data of the single-molecule junction, identify the connection state of the target single-molecule junction.

2. The method for identifying the connection state of a single-molecule junction according to claim 1, wherein The method further includes: Obtaining the first on-off experiment data of the gold-gold junction and marking the first on-off experiment data as positive samples; Obtaining the second on-off experiment data of an unknown structure and determining whether the unknown structure is a single-molecule junction based on the second on-off experiment data; When the unknown structure is a single-molecule junction, marking the second on-off experiment data as negative samples; Constructing a sample data set based on the marked first on-off experiment data and the marked second on-off experiment data, and training a preset model based on the sample data set to obtain the first classification model.

3. The method for identifying the connection state of a single-molecule junction according to claim 2, wherein, The step of determining whether the unknown structure is a single-molecule junction based on the second on-off experiment data includes: Performing feature training on a preset autoencoder based on the first on-off experiment data to obtain a target autoencoder; Determining the mean square error of the second on-off experiment data through the target autoencoder; When the mean square error is greater than a preset threshold, determining that the unknown structure is a single-molecule junction.

4. The method for identifying the connection state of a single-molecule junction according to claim 1, wherein The method further includes: Obtaining the first voltammetry experiment data of the gold-gold junction and marking the first voltammetry experiment data as positive samples; Obtaining the second voltammetry experiment data of an unknown structure; Based on the second voltammetry experiment data and the marked first voltammetry experiment data, performing iterative training on a preset initial classifier to obtain multiple weak classifiers; Combining each of the weak classifiers to obtain a strong classifier, and using the strong classifier as the second classification model.

5. The method for identifying the connection state of a single-molecule junction according to claim 4, wherein, The step of combining each of the weak classifiers to obtain a strong classifier includes: Detecting the number of the weak classifiers; When the number of the weak classifiers reaches a preset number, performing the step of combining each of the weak classifiers to obtain a strong classifier.

6. The method for identifying the connection state of a single-molecule junction according to claim 1, wherein The step of identifying the connection state of the target single-molecule junction based on the on-off data of the gold-gold junction, the on-off data of the single-molecule junction, the voltammetry data of the gold-gold junction, and the voltammetry data of the single-molecule junction includes: Classify the on-off data of the single-molecule junction to obtain on-off data with multiple different conductance levels; Classify the volt-ampere data of the single-molecule junction according to the degree of linearity to obtain volt-ampere data with multiple different degrees of non-linearity; Based on the on-off data of the Au-Au junction, each of the on-off data, the volt-ampere data of the Au-Au junction, and each of the volt-ampere data, identify the connection state of the target single-molecule junction.

7. A single-molecule junction connection state recognition device, characterized in that, The device for identifying the connection state of the single-molecule junction includes: An on-off data analysis module, configured to input the target on-off experimental data of the target single-molecule junction into a first classification model to obtain the on-off data of the Au-Au junction and the on-off data of the single-molecule junction in the target on-off experimental data. The first classification model is trained with the first on-off experimental data of the Au-Au junction as positive samples and the second on-off experimental data of an unknown structure as unlabeled samples. The unknown structure is an Au-Au junction or a single-molecule junction. The Au-Au junction refers to a structure formed by directly connecting two gold electrodes, and the single-molecule junction refers to a structure formed by indirectly connecting two gold electrodes through a single molecule; A volt-ampere data analysis module, configured to input the target volt-ampere experimental data of the target single-molecule junction into a second classification model to obtain the volt-ampere data of the Au-Au junction and the volt-ampere data of the single-molecule junction in the target volt-ampere experimental data. The second classification model is trained with the first volt-ampere experimental data of the Au-Au junction as positive samples and the second volt-ampere experimental data of an unknown structure as unlabeled samples; An identification module, configured to identify the connection state of the target single-molecule junction based on the on-off data of the Au-Au junction, the on-off data of the single-molecule junction, the volt-ampere data of the Au-Au junction, and the volt-ampere data of the single-molecule junction.

8. A single-molecule junction connection state recognition device, characterized in that, The device for identifying the connection state of the single-molecule junction includes: a memory, a processor, and a computer program stored on the memory and executable on the processor. The computer program is configured to implement the steps of the method for identifying the connection state of the single-molecule junction according to any one of claims 1 to 6.

9. A storage medium, characterized in that, The storage medium is a computer-readable storage medium, and a computer program is stored on the computer-readable storage medium. When the computer program is executed by a processor, it implements the steps of the method for identifying the connection state of the single-molecule junction according to any one of claims 1 to 6.

Images