CN117350775A

CN117350775A - Medical surgical robot market current situation demand analysis method and system

Info

Publication number: CN117350775A
Application number: CN202311330805.8A
Authority: CN
Inventors: 姜冠群; 赵玉林
Original assignee: Shandong Zhuoye Medical Technology Co ltd
Current assignee: Shandong Zhuoye Medical Technology Co ltd
Priority date: 2023-10-13
Filing date: 2023-10-13
Publication date: 2024-01-05

Abstract

The invention relates to the technical field of robots, in particular to a medical surgical robot market current situation demand analysis method and system, wherein the method comprises the following steps: acquiring real-time medical surgical robot market current data based on a big data mining technology; carrying out data integration on the current status data of the real-time medical surgical robot market by utilizing a knowledge graph algorithm, and constructing a real-time surgical robot demand database; performing module decomposition on the real-time surgical robot demand database to generate a surgical robot demand data module; performing dynamic user demand analysis on the surgical robot demand data module by utilizing dynamic psychology to obtain dynamic surgical robot demand analysis data; performing feature extraction on the dynamic surgical robot demand analysis data to generate surgical robot demand feature data; the method and the device realize the purpose of providing efficient and accurate market demand trend prediction.

Description

Medical surgical robot market current situation demand analysis method and system

Technical Field

The invention relates to the technical field of robots, in particular to a medical surgical robot market current situation demand analysis method and system.

Background

With the continuous progress of technology and the development of medical industry, medical surgical robots are being widely used in surgical operations as an advanced surgical aid. However, in the medical surgical robot market, it is important to understand market demands and analyze market trends. In the conventional market current demand analysis method, the problem of low demand analysis efficiency and inaccurate analysis result often exists, so that a current demand analysis method and system need to be introduced to better know the demand trend of the medical surgical robot market. By combining a feature engineering algorithm, a cluster analysis method and a cyclic convolution network, a trend prediction convolution model of the surgical robot is constructed, so that the high-efficiency and accurate trend prediction of market demands is realized.

Disclosure of Invention

The invention provides a medical surgical robot market current situation demand analysis method and system for solving at least one of the technical problems.

In order to achieve the above object, the present invention provides a medical surgical robot market current demand analysis method, comprising the following steps:

step S1: acquiring real-time medical surgical robot market current data based on a big data mining technology; carrying out data integration on the current status data of the real-time medical surgical robot market by utilizing a knowledge graph algorithm, and constructing a real-time surgical robot demand database;

Step S2: performing module decomposition on the real-time surgical robot demand database to generate a surgical robot demand data module; performing dynamic user demand analysis on the surgical robot demand data module by utilizing dynamic psychology to obtain dynamic surgical robot demand analysis data;

step S3: performing feature extraction on the dynamic surgical robot demand analysis data to generate surgical robot demand feature data; performing association rule mining on the surgical robot demand feature data to generate surgical robot demand association feature data;

step S4: performing time sequence analysis on the surgical robot demand associated feature data according to a time sequence algorithm to generate surgical robot time sequence demand feature data; market trend prediction is carried out on the time sequence demand characteristic data of the surgical robot by utilizing a random forest algorithm so as to generate demand trend prediction data of the surgical robot;

step S5: and carrying out data visualization on the demand trend prediction data of the surgical robot to generate a demand trend prediction interactive view.

Step S6: and performing expansion convolution on the visual view of demand trend prediction by using a cyclic convolution network, and constructing a surgical robot trend prediction convolution model so as to execute market demand trend prediction operation.

The invention can collect and analyze huge market data through big data mining technology, comprises information such as market scale, growth rate, market share and the like, acquires market current data in real time, can reflect current market dynamics, helps enterprises to know market trends in time, integrates real-time data by utilizing a knowledge-graph algorithm, can construct a comprehensive market database, is convenient for subsequent demand analysis and prediction, integrates real-time data by utilizing the knowledge-graph algorithm, can construct a comprehensive market database, is convenient for subsequent demand analysis and prediction, generates a surgical robot demand data module, is helpful for clearly knowing characteristics and trends of different demands, provides a basis for subsequent analysis, can extract relevant characteristics from original data by characteristic engineering, helps to identify and construct key factors of demands, and is extracted by the characteristics, the method can reduce the dimension of data, improve the efficiency of data processing and analysis, generate the characteristic data of the surgical robot demand, help to understand the user demand and the important characteristics of the demand more deeply, the time sequence analysis can reveal the change mode and trend of the demand along with time, help to understand the periodicity and trend dynamics of the market, the long-term and short-term prediction can be carried out on the demand through the time sequence analysis, help to make reasonable market strategies and decisions, the data visualization can display the complex demand trend prediction result in a graphical mode, can understand and analyze the data more intuitively, the data visualization can display the complex demand trend prediction result in a graphical mode, help to make and analyze the data more intuitively, the data visualization can help to convey the market trend prediction result to business decision makers, help to their corresponding market strategies, the cyclic convolution network can better capture long-term dependency in time sequence data, the prediction accuracy is improved, the receptive field of the model can be increased and richer features are extracted through expansion convolution, the prediction result is further improved, the construction of the trend prediction convolution model of the surgical robot is beneficial to realizing automatic trend prediction, efficient and accurate market demand prediction is provided, and a reference basis is provided for enterprise decision.

Preferably, step S1 comprises the steps of:

step S11: acquiring real-time medical surgical robot market current situation data based on a big data mining technology, wherein the real-time medical surgical robot market current situation data comprises market scale data, robot product information data, robot field data, robot market share data and clinical use case data;

step S12: carrying out relation extraction on the market current data of the real-time medical surgical robot by using a knowledge graph algorithm to generate market current node data;

step S13: and (3) integrating the data of the market current node to construct a real-time surgical robot demand database.

The invention is helpful for knowing the size and the growing trend of the medical surgical robot market by acquiring the real-time market scale data. This provides important indicators for enterprise decision makers and investors to assess market potential, including product information for various medical surgical robots, such as technical features, functions, performance indicators, and the like. The data can help enterprises to know the difference and the competitive advantage among different products, provide references for product development, positioning and marketing, acquire real-time robot field data and know the application condition and market requirements of the medical surgical robot in different medical fields. The method is beneficial to enterprises to grasp market trends, discover new application fields and optimize product research and development directions, and the relative status and competition conditions of different manufacturers and products in the market can be provided by acquiring market share data of the medical surgical robot in real time. This is of great significance to enterprises in developing marketing strategies, finding partners, and assessing their own location in the marketplace, knowledge-graph algorithms can identify and extract relationships between entities from raw data, such as associations between products and technologies, relationships between market size and market share, and so forth. This facilitates the construction of a data model with a structured representation, by which real-time medical surgical robot market-state data can be transformed into a set of node data, nodes representing different entities, edges representing the relationships between them. The data representation form can better display and analyze the market status, is convenient for subsequent demand analysis and prediction, integrates market status node data, and can integrate information of different data sources to construct a comprehensive surgical robot demand database. The method and the system enable various market data to be shared and uniformly managed, facilitate subsequent demand analysis, and integrate and analyze information of different data sources under a uniform view through data integration. This helps to discover correlations and trends between different data, providing more comprehensive market insight and decision support.

Preferably, step S2 comprises the steps of:

step S21: carrying out structural decomposition on a real-time surgical robot demand database to generate a multi-stage surgical robot demand data structure;

step S22: performing module decomposition on the multi-stage surgical robot demand data structure to generate a surgical robot demand data module;

step S23: performing behavior analysis on the surgical robot demand data module by using behavior psychology to generate market behavior habit data;

step S24: and carrying out dynamic user demand analysis on the market behavior habit data to obtain dynamic surgical robot demand analysis data.

The invention can organize and classify the complex demand information by carrying out structural decomposition on the demand database to form a hierarchical structure. This helps to clearly represent the different levels of demand relationships and dependencies, facilitating subsequent analysis and understanding, by structurally decomposing, generating a multi-level surgical robot demand data structure, subdividing the demand into different levels and modules. This facilitates finer granularity of management and analysis of demand, provides more detailed demand information, and further breaks the multi-level data structure into different modules, each representing a particular demand field or function. The organization mode facilitates independent management and analysis of different modules, provides more flexible data operation and query capability, and can refine requirements to more specific functional and characteristic levels through module decomposition. The method is helpful for deeply understanding details and differences of user demands, provides clear guidance and reference for product design and development, and can analyze behavioral patterns and psychological factors of users when the surgical robots are actually used by the users through a behavioral psychology method. The method is favorable for knowing the preference, habit and behavior motivation of the user, provides a targeted strategy for product design and marketing, and can generate market behavior habit data, namely behavior habit and behavior characteristics of the user in the field of surgical robots through behavior analysis. The data can be used for market positioning, user portrayal and product customization, help enterprises predict market demands and formulate corresponding strategies, and can know dynamic changes and trends of the user demands through analysis of market behavior habit data. This helps the enterprise follow up market changes, adjust product strategies and innovation directions in time, remain competitive, and dynamic user demand analysis can provide insight into market demand, including user expectations for new functions and technologies, pain points, demand priorities, and the like. The data can be used for product planning, function optimization and user experience improvement, and the actual requirements of users are met.

Preferably, step S3 comprises the steps of:

step S31: carrying out feature extraction on the dynamic surgical robot demand analysis data by using a feature engineering method to generate surgical robot demand feature data;

step S32: performing dimension reduction processing on the demand characteristic data of the surgical robot by using a principal component analysis method, so as to generate demand characteristic dimension reduction data;

step S33: clustering and dividing the demand characteristic dimension reduction data by using a clustering analysis method to generate demand characteristic clustering data;

step S34: performing association rule mining on the demand feature cluster data to generate a demand feature association set,

step S35: and screening the feature rule of the requirement feature association set to generate the surgical robot requirement association feature data.

The invention extracts the characteristics which are significant for demand analysis from the original data through the characteristic engineering. By applying various feature extraction techniques, such as statistical features, frequency features, time series features, etc., the most representative and differentiated feature sets can be selected and constructed. This will help reduce the dimensionality of the data, remove redundant information, and provide more informative data for subsequent analysis, principal Component Analysis (PCA) being a commonly used method of dimension reduction. It maps the high-dimensional feature space to the low-dimensional space by linear transformation while preserving as much variance information of the original data as possible. By reducing the data dimension, the complexity and computational complexity of the data can be reduced while maintaining important information on the overall structure of the data for subsequent clustering and mining analysis, which is an unsupervised learning method that classifies similar data objects into the same category to form groups. By performing cluster analysis on demand feature dimension reduction data, similar demands can be clustered together and categorized into one class. The method is helpful for finding the similarity and the difference between different requirements, provides a basis for subsequent association rule mining, and is a method for searching interesting relations among items in the data set. And applying an association rule mining algorithm to the demand feature cluster data to reveal the association between different features and generate an association set with measurement indexes such as support, confidence and the like. This helps to find dependencies and dependencies between demand features, providing a basis for further feature rule screening, in which there may be a large number of association rules in the demand feature association set, some of which may be irrelevant or non-representative. Through feature rule screening, those most significant and beneficial feature association rules can be selected. This helps to extract key demand-related features, providing important reference information for product design, market location, and user demand analysis.

Preferably, step S32 comprises the steps of:

step S321: data standardization is carried out on the demand characteristic data of the surgical robot, and demand characteristic standardization data are generated;

step S322: performing feature matrix decomposition on the demand feature standardized data to generate demand feature matrix data;

step S323: vector mapping is carried out on the demand feature matrix data by using a principal component analysis method, and the demand feature matrix vector data is generated;

step S324: and performing dimension reduction processing on the demand feature matrix vector data so as to generate demand feature dimension reduction data.

The invention converts the characteristic data with different scales and ranges into the data with uniform scales through data standardization. By normalizing the surgical robot demand feature data, the numerical differences between different features can be eliminated, so that the feature data is comparable and interpretable. This helps to avoid bias or effects due to differences in feature values during subsequent analysis, principal Component Analysis (PCA) is a commonly used feature extraction and dimension reduction method. The demand feature matrix data can be converted into a set of linearly independent vectors, each representing a principal component, by principal component analysis. This will help capture and represent the primary information and variance distribution in the demand signature data while reducing the dimensionality of the data, the purpose of the dimension reduction process being to reduce the dimensionality of the data while maintaining as much information of the original data as possible. By dimension reduction of the demand feature matrix vector data, high-dimension data can be mapped to a low-dimension space, so that the calculation efficiency and the visualization capability are improved. The method is helpful for reducing the complexity of the data, removing redundant information of the data, extracting the most representative characteristics and providing a basis for subsequent demand analysis and mining.

Preferably, step S4 comprises the steps of:

step S41: performing time sequence analysis on the surgical robot demand associated feature data according to a time sequence algorithm to generate surgical robot time sequence demand feature data;

step S42: performing time sequence demand weight calculation on the time sequence demand characteristic data of the surgical robot by using a time sequence demand weight calculation formula so as to generate time sequence demand weight characteristic data;

step S43: trend analysis is carried out on the time sequence demand weight characteristic data by utilizing a random forest algorithm, and time sequence demand trend analysis data is generated;

step S44: performing periodic trend prediction on the time sequence demand trend analysis data based on a linear regression method to generate time sequence demand periodic trend prediction data;

step S45: carrying out market trend prediction on the time sequence demand cycle trend prediction data by using a market demand trend prediction calculation formula so as to generate surgical robot demand trend prediction data;

according to the invention, through time sequence analysis, the change trend of the surgical robot demand data along with time can be researched and understood. This will help to discover seasonal, trending, and periodic patterns of demand data and extract time-dependent demand characteristic data. The generated surgical robot time sequence demand characteristic data can be used as input of subsequent steps for further analysis and prediction, and the importance and contribution degree of the surgical robot time sequence demand characteristic at different time points can be evaluated by calculating time sequence demand weight. This may be calculated based on a specific time series demand weight calculation formula, based on the change in demand characteristic data and the weight assignment. The generated time sequence demand weight characteristic data can help to know the relative importance and influence degree of the demand characteristics in different time periods, and the trend analysis is carried out on the time sequence demand weight characteristic data through a random forest algorithm, so that the long-term trend and the change mode of the demand characteristic data can be explored. Random forests are a machine learning algorithm that can provide trend analysis of demand features by building multiple decision tree models for ensemble learning. The generated time sequence demand trend analysis data can help to be informed of the overall trend of the demand, the direction and the amplitude of long-term change, and the time sequence demand trend analysis data is used for carrying out periodical trend prediction by using a linear regression method. Linear regression is a statistical analysis method for establishing relationships between variables, and by selecting a best fit line, the periodic variation of demand trends can be predicted. The generated time sequence demand period trend prediction data can help to know the periodic mode of demand change and provide prediction of demand in a certain time range in the future, and the time sequence demand period trend prediction data can be mapped to trend prediction of market demand through a market demand trend prediction calculation formula. This may calculate demand trend prediction data for the surgical robot based on market analysis and related metrics. The generated surgical robot demand trend prediction data can guide decision making, help predict the trend of future demand changes, and plan and adjust related design, development and marketing strategies.

Preferably, the time-series demand weight calculation formula in step S42 is specifically:

wherein W is a time sequence demand weight value, i is an ith market demand analysis feature point, n is the total number of the market demand analysis feature points, and alpha _i Analyzing feature weights, p, for the ith market demand _i Analyzing the market scale value, p, of the feature points for the ith market demand _(i,t) Analyzing the market scale value of the feature point at the t time point for the ith market demand, wherein h is the market demand period value, j is the jth medical robot type, m is the total amount of the medical robot types, and w _j The j-th medical robot type occupies the weight of the market scale, w _(j,t) The market scale weight of the jth medical robot at the t-th time point is given.

The invention is realized byThe relative change rate of the market demand analysis feature points is calculated. The increasing or decreasing trend of the feature point at different time points can be measured by calculating the natural logarithm of the ratio of the market scale value of the feature point to the market scale value of the feature point at the specific time point. A larger ratio indicates a larger relative change, while a smaller ratio indicates a smaller relative change, and the summation calculates a weighted sum of the relative change rates of all feature points. By weighting and summing all the characteristic points, the importance of different characteristic points can be comprehensively considered, the overall change trend of market demands can be captured, The volatility of the market demand is calculated. The degree of fluctuation of the market demand can be measured by calculating the square of the difference between the market size value of the feature point and the market size value of the specific time point. Then, multiplying the market scale weight of the medical robot type with the market scale weight of the specific time point, the importance of different robot types and different time points can be considered, the overall volatility of the market demand can be calculated by weighting and summing all the characteristic points and all the robot types and applying the limit operation h-0, and the method comprises the following steps of p _(i,t) -p _i ² And (3) carrying out square root operation on the fluctuation calculated in the third step, and obtaining the standard deviation of market demands. The standard deviation measures the overall variation amplitude of the market demand, a larger standard deviation indicates that the market demand has larger fluctuation, and a smaller standard deviation indicates that the market demand has smaller fluctuation.

Preferably, the market demand trend prediction calculation formula in step S45 is specifically:

wherein M is a market demand trend prediction result value, B is a market penetration rate, D is a market share occupied amount, E is a globalization influence factor, F is the number of medical robot mechanisms, G is a medical robot price increase rate, L is a market product annual trend index, L is a consumer confidence index, M is a medical surgical robot average service life, O is a medical surgical robot technology iterative update period, R is a market price sensitivity factor, W is a market economic environment variable, X is a historical market share, and Y is a market demand influence degree factor.

The present invention calculates the natural logarithm of the market penetration rate B by lnB. By taking the logarithm, the exponentially increasing market penetration rate can be converted into a linearly increasing form, which is easier to compare and analyze, and the product of the market share occupancy D and the globalization influence factor E and the number of medical robots F is calculated by d×tan (e×f). By combining market share with the product of globalization effect and number of institutions, the increasing trend of market share and the effect of globalization and institution factor can be considered byThe product of the market price sensitivity factor R and the market economic environment variable W, the historical market share X and the market demand influence degree factor Y is calculated, and the result of the limit operation W-0 is calculated. This calculation takes into account the combined impact of market price sensitivity, economic environment, historical market share and market demand influencing factors on market demand, and obtains the market demand trend prediction result value M by dividing the weighted sum of the results of the previous calculations by the weighted sum. This value represents a prediction of market demand trend based on given variables and factors. A larger predictor value indicates that market demand may increase, while a smaller predictor value indicates that market demand may decrease.

Preferably, step S5 comprises the steps of:

step S51: carrying out frequent item set mining on the demand trend prediction data of the surgical robot by using a deep learning algorithm to generate a demand trend prediction feature vector;

step S52: performing data visualization on the demand trend prediction feature vector to generate a demand trend prediction visual view;

step S53: and carrying out interactive processing on the visual view of the demand trend prediction by using a JavaScript library to generate the interactive view of the demand trend prediction.

According to the invention, frequent item set mining is carried out on the demand trend prediction data through a deep learning algorithm, so that frequent modes and association rules in the demand trend can be found. This may help identify correlations and correlations between different demand features, generating demand trend predictive feature vectors. These feature vectors may provide a more detailed and comprehensive description that helps understand patterns and rules in demand trends, and by visualizing data on demand trend predictive feature vectors, abstract data may be transformed into visualized graphs and charts. This helps to intuitively demonstrate the nature and change in demand trends, making the data easier to understand and interpret. The generated visual view of the demand trend prediction can provide a global view angle to help find out the mode, the abnormality and the change of the trend in the trend, and more exploration and interaction functions can be provided by utilizing the JavaScript library to carry out interactive processing on the visual view of the demand trend prediction, so that a user can control and analyze data according to own demands and interests. This includes zoom, scroll, swipe, interactive annotation, etc. functions that make the visual view of demand trend prediction more flexible and operable. The generated demand trend predictive interactive view can help users freely explore data and discover deeper information and insights.

Preferably, step S6 comprises the steps of:

step S61: carrying out convolution pretreatment on the visual view of demand trend prediction by using a cyclic convolution network to generate a demand trend prediction convolution sample set;

step S62: performing convolution data cutting on the demand trend prediction convolution sample set to generate a demand trend prediction convolution sequence;

step S63: performing expansion convolution on the demand trend prediction convolution sequence by using a cavity convolution algorithm to generate a demand trend prediction convolution network;

step S64: carrying out space pyramid pooling multi-layer sampling on the demand trend prediction convolution network to generate a demand trend prediction convolution feature map;

step S65: and stacking and integrating modeling is carried out on the demand trend prediction convolution feature map by utilizing a combined classifier algorithm, and a surgical robot demand trend prediction convolution model is constructed so as to execute market demand trend prediction operation.

According to the invention, the feature information in the image can be extracted by carrying out convolution preprocessing on the visual view of demand trend prediction by applying a cyclic convolution network. This may translate abstract visual data into a representation of features with a higher level to better capture subtle changes and patterns in demand trends. The generated demand trend prediction convolution sample set may provide richer and more valuable data for further analysis and modeling, and the demand trend prediction convolution sample set is data cut into a plurality of sequential sequence data. This has the advantage that the timing information of the data can be preserved, enabling the model to capture the time dependence in demand trends. The generated demand trend predictive convolution sequence can be used for further processing and modeling. By applying the hole convolution algorithm, the receptive field of the convolution kernel can be enlarged without increasing parameters. This allows for better capture of long-range dependencies and global features in demand trends. The generated demand trend prediction convolution network has stronger perceptibility and representation capability, is beneficial to improving the prediction accuracy and generalization capability of the model, and can carry out pooling operation on the feature map from different scales by applying a spatial pyramid pooling multi-layer sampling method so as to capture the features of different spatial levels in the data. This may provide a richer and diversified representation of the features, helping to increase the flexibility and generalization ability of the model. The generated demand trend predictive convolution feature map can be used for subsequent feature extraction and modeling, and multiple classifiers can be combined by using a combined classifier algorithm to improve the performance and robustness of the model. This includes integration algorithms such as random forests, gradient lifting trees, etc. Constructing a surgical robot demand trend predictive convolution model can help predict market demand trend and make corresponding decisions. The model utilizes rich information extracted from the convolution feature map and a plurality of classifiers which are stacked and integrated to improve prediction accuracy and stability.

In this specification, there is provided a medical surgical robot market-state demand analysis system including:

the data integration module is used for acquiring real-time medical surgical robot market current situation data based on a big data mining technology; carrying out data integration on the current status data of the real-time medical surgical robot market by utilizing a knowledge graph algorithm, and constructing a real-time surgical robot demand database;

the demand analysis module is used for carrying out module decomposition on the real-time surgical robot demand database so as to generate a surgical robot demand data module; performing dynamic user demand analysis on the surgical robot demand data module by utilizing dynamic psychology to obtain dynamic surgical robot demand analysis data;

the feature engineering module is used for carrying out feature extraction on the dynamic operation robot demand analysis data by using a feature engineering method to generate operation robot demand feature data; performing association rule mining on the surgical robot demand feature data by using a cluster analysis method to generate the surgical robot demand association feature data;

the trend prediction module is used for carrying out time sequence analysis on the surgical robot demand associated feature data according to a time sequence algorithm to generate surgical robot time sequence demand feature data; market trend prediction is carried out on the time sequence demand characteristic data of the surgical robot by utilizing a random forest algorithm so as to generate demand trend prediction data of the surgical robot;

And the data visualization module is used for carrying out data visualization on the demand trend prediction data of the surgical robot by using a deep learning algorithm to generate a demand trend prediction interactive view.

And the convolution model module is used for performing expansion convolution on the visual view of the demand trend prediction by using a circular convolution network, and constructing a surgical robot trend prediction convolution model so as to execute market demand trend prediction operation.

According to the invention, through data integration, the system can acquire comprehensive real-time market current data, a unified database is constructed, a basis is provided for subsequent demand analysis, and through a modularized analysis and dynamic psychological method, the system can deeply understand the demands of users, acquire psychological factors and behavioral characteristics related to the demands, and provide a basis for subsequent characteristic engineering. Feature engineering can reduce the dimensionality of data and extract the most relevant features, association rule mining can reveal hidden association modes in demand data, and a foundation is laid for demand trend prediction. According to the time sequence algorithm, time sequence analysis is carried out on the surgical robot demand associated feature data, trends and periodic modes in the data are explored, market trend prediction is carried out on the time sequence demand feature data by using a random forest algorithm, future demand trend is predicted, and a trend prediction module can help to predict the development trend of the surgical robot demand and provide decision support and planning prediction. The data visualization module can display the predicted demand trend in an intuitive and easy-to-understand manner, providing more intuitive data insight and interactive analysis. The convolution model can capture long-range dependence and global features in the demand trend through expansion convolution, so that the prediction accuracy and robustness of the model are improved.

Drawings

FIG. 1 is a schematic flow chart of steps of a medical surgical robot market current demand analysis method and system according to the present invention;

FIG. 2 is a detailed implementation step flow diagram of step S1;

FIG. 3 is a detailed implementation step flow diagram of step S2;

fig. 4 is a detailed implementation step flow diagram of step S3.

Detailed Description

It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.

The application example provides a medical surgical robot market current situation demand analysis method and system. The execution main body of the medical surgical robot market current situation demand analysis method and the system comprises, but is not limited to, the system: mechanical devices, data processing platforms, cloud server nodes, network uploading devices, etc. may be considered general purpose computing nodes of the present application, including but not limited to: at least one of an audio image management system, an information management system and a cloud data management system.

Referring to fig. 1 to 4, the present invention provides a medical surgical robot market current demand analysis method, which includes the steps of:

In the embodiment of the present invention, as described with reference to fig. 1, a schematic flow chart of steps of a method and a system for analyzing market status requirements of a medical surgical robot according to the present invention is provided, and in this example, the steps of the method for analyzing market status requirements of a medical surgical robot include:

in this embodiment, a data source that can be used to obtain real-time data, such as an online transaction platform, an industry report, a social media platform, medical institution data, etc., is determined, and a big data mining technique is used to capture and extract the required data. This may include crawling data from web pages using web crawler technology, or connecting to related platforms and databases using API interfaces to obtain data, cleaning and preprocessing data collected from different data sources, ensuring quality and consistency of the data. And then integrating and correlating the data by utilizing a knowledge graph algorithm, constructing a real-time surgical robot demand database, cleaning and preprocessing the data collected from different data sources, and ensuring the quality and consistency of the data. And then integrating and correlating the data by utilizing a knowledge graph algorithm to construct a real-time surgical robot demand database, defining entity types, relation types and attributes according to the knowledge field and the data attributes of the surgical robot market when the knowledge graph is constructed, mapping the data to nodes and edges in the graph database, and establishing a database management system to ensure the availability of the real-time surgical robot demand database and the timely updating of the data. Designing appropriate data storage structures and indexes to improve data access and query efficiency

in this embodiment, the real-time surgical robot demand database is decomposed in modules, and the data in the database are classified and generalized according to different modules. The modules can be divided according to the characteristics, the purposes or the fields of the data, such as a surgical robot functional module, a user evaluation module, a market trend module and the like, the real-time surgical robot demand database is subjected to module decomposition, and the data in the database are classified and generalized according to different modules. The modules can be divided according to the characteristics, the purposes or the fields of the data, such as a surgical robot functional module, a user evaluation module, a market trend module and the like, and the dynamic psychological method is utilized to analyze the dynamic user requirements of the surgical robot requirement data module. Dynamic psychology is a discipline for researching dynamic changes of psychological activities in time, and can be used for understanding how user demands evolve and change with time, collecting actual demand data of users according to a designed data collection method, and carrying out data arrangement and analysis. The data may be processed and parsed using appropriate statistical methods, text analysis techniques, emotion analysis, etc. to obtain dynamic surgical robot demand analysis data.

in this embodiment, feature extraction is performed on dynamic surgical robot demand analysis data. Feature engineering refers to converting original data into more meaningful features by means of conversion, combination or selection according to the attributes and characteristics of the data so as to better represent the meaning and features of the data. Various characteristics related to the requirements of the surgical robot can be extracted from the original data, and the selected characteristics are appropriately encoded and converted into a form suitable for cluster analysis. The common coding method comprises single-heat coding, label coding and the like, a proper coding mode is selected according to specific conditions, and the clustering analysis method is used for processing the characteristic data required by the surgical robot. Cluster analysis is a technique that organizes similar data points into clusters, aiming at finding intrinsic structures and patterns in the data. Various clustering algorithms, such as K-means clustering, hierarchical clustering and the like, can be used, a proper algorithm is selected according to specific requirements, and correlation and association rules among different features are found out by using an association rule mining method in clusters obtained through cluster analysis. Correlation rule mining can reveal correlation, dependency and influence relation among variables, so that correlation features in surgical robot requirements are better understood, and the correlation features are combined into surgical robot requirement correlation feature data according to the result of the correlation rule mining.

in this embodiment, the surgical robot demand-related feature data is ordered according to a time sequence, so that the data points in the data set are arranged according to the time sequence, and a proper time sequence model is selected according to the target and the data characteristics of time sequence analysis. Common time series models include ARIMA model, seasonal ARIMA model (SARIMA), exponential smoothing, and the like. The method comprises the steps of selecting a proper model, determining according to the seasonal, trending and noise degree of data, dividing a time sequence data set into a training set and a testing set according to a certain proportion, fitting the selected time sequence model by using the data of the training set to obtain parameters and fitting results of the model, and predicting the surgical robot demand in a future period by using the fitted time sequence model. The future demand trend can be predicted through the parameters and the historical data of the model, the demand trend prediction data of the surgical robot are generated, and the market trend is predicted on the generated time demand characteristic data of the surgical robot by using a random forest algorithm. Random forests are an integrated learning algorithm that predicts by combining multiple decision trees. It can take into account the importance and interaction of the individual features, thereby improving the accuracy of the predictions.

In this embodiment, the data is converted into a format suitable for visualization according to the requirements of the visualization library. It is often convenient to store data in tabular form, and common data processing tools (such as the pandas library in Python) may be used to perform data format conversion, and appropriate chart types and interface layouts may be designed according to the characteristics and visualization purposes of the data. For example, a line graph, area graph, or bar graph may be selected to show trends in surgical robot demand, and to add functionality to interact with time (e.g., zoom in and out, scroll view, etc.), enabling the user to freely explore the data, using a selected visualization library to map the predicted data into an interactive chart. According to the time sequence characteristics of the data, taking the time stamp as an abscissa and the demand predicted value as an ordinate, drawing corresponding data points on a chart, and adding the interaction function according to the function and the demand of the library. This may include zoom-in and zoom-out functions, hover cues for data points (e.g., displaying specific values), time range selectors, and the like. These interactive functions enable the user to view different details of the data and interact with the data as desired.

In this embodiment, a suitable cyclic convolution network (e.g., LSTM, GRU, etc.) is selected to construct the surgical robot demand trend prediction model. The cyclic convolution network can capture the time sequence dependency of time sequence data and perform well on the prediction problem, and the expansion convolution operation is performed on the demand trend prediction visual view so as to extract richer and more representative features. The expansion convolution (Dilated Convolution) is a convolution operation, a certain degree of holes (position) are introduced and the receptive field is enlarged on the basis of the traditional convolution operation, so that long-term dependence in time sequence data can be better captured, and a constructed cyclic convolution network model is trained by using a preprocessed data set. In the training process, the data set is divided into a training set, a verification set and a test set, the training set is used for optimizing model parameters, the verification set is used for monitoring the performance of the model and adjusting the performance, finally the test set is used for evaluating the generalization capability of the model, and the training and evaluating surgical robot trend prediction convolution model is used for predicting future market demand trend. And according to the historical data and other influencing factors, inputting the historical data and other influencing factors into a model for prediction, and generating a demand trend prediction result.

In this embodiment, as described with reference to fig. 2, a detailed implementation step flow diagram of the step S1 is described, and in this embodiment, the detailed implementation step of the step S1 includes:

In this embodiment, real-time medical surgical robot market status data is extracted from the selected data sources using a large data mining technique, such as a web crawler. According to the required data types (market scale, product information, field data, market share, clinical use cases and the like), a corresponding grabbing strategy is formulated, a crawler program is written to acquire data, and the grabbed data is cleaned, including duplicate data removal, missing value processing, unified format and the like. The cleaned data is easier to carry out subsequent processing and analysis, and a structural model of the knowledge graph is designed according to the characteristics and the relation of the current market data of the real-time medical surgical robot. Determining nodes (such as products, fields, market shares and the like) and relations among the nodes (such as the corresponding fields of the products, the relations between the market shares and the products and the like), and performing entity identification and relation extraction on the acquired real-time data by utilizing technologies such as Natural Language Processing (NLP), machine learning and the like. Extracting entities (such as products, fields, market shares and the like) and the relation among the entities from the captured data, constructing nodes and edges of a knowledge graph, adding the nodes and the edges into the knowledge graph according to the results of entity identification and relation extraction to form a complete market current knowledge graph, preprocessing and standardizing the market current data extracted from the knowledge graph, and ensuring the consistency and comparability of the data. This includes removing noise, unifying units, processing missing values, etc., integrating the preprocessed market-present data with other related data sets, such as historical sales data, market trend data, etc. The information of a plurality of data sources can be integrated through data integration, so that more comprehensive and accurate real-time operation robot demand data can be obtained, the integrated real-time operation robot demand data is stored in a database,

In this embodiment, as described with reference to fig. 3, a detailed implementation step flow diagram of the step S2 is shown, and in this embodiment, the detailed implementation step of the step S1 includes:

In this embodiment, a real-time surgical robot demand database is comprehensively analyzed, information such as data fields and table relationships in the database is known, demand data is structurally decomposed according to an analysis result of the database, and the data is organized and classified according to different levels. For example, the system can be divided according to dimensions such as product types, application fields, market regions and the like to form a multi-stage surgical robot demand data structure, the association relationship among all the layers is determined, and connection among the layers is established so as to be capable of performing cross-layer data query and analysis in the subsequent analysis process, and the multi-stage surgical robot demand data structure is further decomposed into different modules or fields according to the multi-stage surgical robot demand data structure. For example, the hardware requirements, software requirements, operator interface requirements, etc. of the surgical robot may be divided into different modules, and for each required module, corresponding data may be extracted from the real-time surgical robot requirements database. According to the characteristics of the modules, proper query and filtering conditions are selected to obtain required data, and the data extracted from the database are organized and integrated according to the modules to form the surgical robot demand data module. The consistency and accuracy of the data are ensured, and the behaviors of the user in the process of selecting, using and meeting the requirements of the surgical robot are analyzed by using a behavioral psychology theory and a behavioral psychology method according to the surgical robot requirement data module. The method can analyze the user preference, decision making process, purchasing behavior and other aspects, and utilizes the data mining technology and the statistical analysis method to mine and analyze the surgical robot demand data module. The system can mine information such as behavior patterns, demand change trends and the like of users, and according to results of behavior analysis, the obtained insights and conclusions are arranged into market behavior habit data, wherein the market behavior habit data comprise user preference data, decision path data, purchasing habit data and the like, and dynamic user demand analysis is performed by combining the market behavior habit data and real-time surgical robot demand data so as to obtain dynamic surgical robot demand analysis data.

In this embodiment, as described with reference to fig. 3, a detailed implementation step flow diagram of step S3 is described, and in this embodiment, the detailed implementation step of step S1 includes:

In this embodiment, the characteristics associated with the surgical robot requirements are selected based on the goals of the requirements analysis. The importance of features can be determined using domain knowledge, correlation analysis, statistical methods, etc., and new features can be created by combining, deriving, or converting the original features as needed for demand analysis. For example, differences, ratios between features, or statistical features extracted by an aggregation operation may be calculated, and the surgical robot demand feature data is normalized so that each feature has the same scale. Common normalization methods include Z-score normalization and Min-Max normalization, and the normalized demand characteristic data is subjected to dimension reduction processing by using a Principal Component Analysis (PCA) method. PCA converts the original features into a group of main components which are not related to each other through linear transformation so as to capture main change information in the data, and determines the number of reserved dimensions according to the interpretation variance proportion of the main components after dimension reduction. The principal components with higher interpretation variance ratio are generally selected to preserve the variability of most of the original features, and the number of preserved dimensions is determined according to the interpretation variance ratio of the principal components after dimension reduction. The principal components with higher variance ratio are generally selected for interpretation to preserve the variability of most of the original features, and the number of clusters is determined based on the requirements and the goals of the data analysis. The sample can be divided into different clusters by performing cluster analysis on the demand feature dimension reduction data through a visualization method (such as elbow rule and contour coefficient) or field knowledge selection. Each cluster represents the surgical robot demands with similar demand characteristics, parameters of association rule mining, such as support degree, confidence degree and the like, are set according to the target of demand analysis, and an association rule mining algorithm is applied to mine frequent item sets and association rules from demand characteristic cluster data. The frequent item sets represent feature combinations which frequently and simultaneously occur, the association rules represent the association between the features, and the rules are screened and filtered according to the support degree, the confidence degree and other evaluation indexes of the demand feature association sets. Rules with high confidence and support are selected as the associated features of the surgical robot requirements, the screened feature rules are interpreted and analyzed, the associated relationship between the features is understood, and useful insights and insights about the requirements are provided.

In this embodiment, step S32 includes the steps of:

In this embodiment, the data normalization is to ensure that the dimensions of the different features are consistent so that they can be compared and analyzed. The commonly used standardization method comprises Z-score standardization and Min-Max standardization, a proper standardization method is applied to the demand characteristic data of the operation robot to obtain the demand characteristic standardization data, a characteristic matrix is a matrix containing a plurality of characteristics, each row represents one sample, each column represents one characteristic, matrix decomposition is carried out on the demand characteristic standardization data, and a Principal Component Analysis (PCA) method and the like can be adopted. The PCA converts original features into a group of main components which are not related to each other through linear transformation so as to capture main change information in data, performs feature matrix decomposition on demand feature standardized data to obtain demand feature matrix data, and the main component analysis (PCA) maps the feature matrix data into a group of main component vectors which are not related to each other through linear transformation so as to capture the main change information in the data, and applies PCA algorithm to analyze and transform the demand feature matrix data. And obtaining the required feature matrix vector data, wherein each sample is represented by a group of principal component vectors, and dimension reduction is to reduce the dimension of a data set on the premise of keeping main information of the data so as to better visualize, analyze and understand the data, apply a dimension reduction method to the required feature matrix vector data, and select proper dimension reduction methods and parameters by common methods including PCA, linear Discriminant Analysis (LDA) and the like so as to reduce the required feature matrix vector data to a lower dimension. The dimension after dimension reduction can be determined according to the explained variance ratio, the reserved feature quantity or domain knowledge.

In this embodiment, step S4 includes the following steps:

In this embodiment, the requirement related feature data of the surgical robot is collected, including time points and values of related features, and time sequence analysis is performed on the data, and common time sequence analysis methods, such as a smoothing method, seasonal decomposition, an exponential smoothing method, an ARIMA model, and the like, may be adopted, and according to the selected time sequence method, time sequence analysis is performed on the requirement related feature data of the surgical robot, and time sequence requirement feature data of the surgical robot is generated, and a time sequence requirement weight calculation formula is defined, which is used for calculating the weight of the time sequence requirement feature data of the surgical robot. The specific formula can be defined according to the demand characteristics and the business demands, the time sequence demand characteristic data of the surgical robot are calculated according to the defined formula to obtain the demand weight of each time point, the calculated time sequence demand weight is associated with the corresponding time point to generate time sequence demand weight characteristic data, for example, when 7 to 8 months are taken, partial college graduates can select myopia correction surgery, at the moment, the market share weight ratio of the auxiliary medical robot related to the myopia correction surgery on the market can be larger, the data can be better weighted through the time sequence demand weight characteristic data, and in subsequent calculation, the market demand prediction result is more accurate. The random forest is an integrated learning algorithm, can be used for regression and classification problems, takes time sequence demand weight characteristic data as input, establishes a random forest model, generates time sequence demand trend analysis data according to the result of the random forest algorithm, determines trend states of each time point, establishes a model by using a linear regression method, takes time as an independent variable, takes the time sequence demand trend analysis data as a dependent variable, fits and predicts the established linear regression model to obtain a periodic trend prediction result of the time sequence demand trend analysis data, generates time sequence demand periodic trend prediction data according to the prediction result of the linear regression model, reflects periodic trends in the time sequence data, and defines a market demand trend prediction calculation formula which is used for carrying out market trend prediction according to the time sequence demand periodic trend prediction data. The specific formula can be defined according to market demand analysis and business demand, the time sequence demand period trend prediction data is used for calculation according to the defined formula to obtain a market trend prediction result, the market trend prediction result is associated with a corresponding time point, and the surgical robot demand trend prediction data is generated.

In this embodiment, the time-series demand weight calculation formula in step S42 is specifically:

The invention is realized byThe relative change rate of the market demand analysis feature points is calculated. The increasing or decreasing trend of the feature point at different time points can be measured by calculating the natural logarithm of the ratio of the market scale value of the feature point to the market scale value of the feature point at the specific time point. A larger ratio indicates a larger relative change, while a smaller ratio indicates a smaller relative change, and the summation calculates a weighted sum of the relative change rates of all feature points. By weighting and summing all the characteristic points, the importance of different characteristic points can be comprehensively considered, the overall change trend of market demands can be captured, The volatility of the market demand is calculated. By calculating market scale values of feature points with specific time pointsThe square of the difference between market scale values can measure the degree of fluctuation of market demand. Then, multiplying the market scale weight of the medical robot type with the market scale weight of the specific time point, the importance of different robot types and different time points can be considered, the overall volatility of the market demand can be calculated by weighting and summing all the characteristic points and all the robot types and applying the limit operation h-0, and the method comprises the following steps of p _(i,t) -p _i ² And (3) carrying out square root operation on the fluctuation calculated in the third step, and obtaining the standard deviation of market demands. The standard deviation measures the overall variation amplitude of the market demand, a larger standard deviation indicates that the market demand has larger fluctuation, and a smaller standard deviation indicates that the market demand has smaller fluctuation.

In this embodiment, the market demand trend prediction calculation formula in step S45 is specifically:

In this embodiment, step S5 includes the following steps:

In this embodiment, a deep learning algorithm is used to perform frequent item set mining. Common deep learning algorithms include association rule mining, apriori algorithm, FP-growth algorithm, and the like. These algorithms can help us find frequent item sets in the data, i.e. feature combinations that frequently occur simultaneously in the time series data, and extract the frequent item sets as feature vectors for demand trend prediction. Each frequent item set can be regarded as a feature, the feature vector is composed of a plurality of frequent item sets, whether the corresponding feature appears or not is represented by the appearance or not of each frequent item set, the feature vector dimension is predicted according to the requirement trend, a proper data visualization method such as a line graph, a bar graph, a scatter graph and the like is selected to show the distribution, the variation trend and the mutual relation of the feature, a proper axis label, a legend, color coding and the like are added in the visualized view to enhance the readability and the interpretation, the visualized view of the plurality of features can be combined and displayed according to the requirement, such as the representation in a sub-graph or parallel graph mode, interactive elements such as sliding bars, drop-down menus, check boxes and the like are added in the view to allow a user to customize the displayed feature, the time range and the like, the interactive function with the view is realized, such as changing the displayed data through the interactive elements, zooming in the view, clicking the detailed information and the like, the interactive view is predicted according to the requirement trend, the use scene and the requirement of the visual interactive view is interacted, and the user can be ensured to conveniently explored and the data are analyzed according to the specific functional requirement.

In this embodiment, step S6 includes the following steps:

In this embodiment, the visual view data is convolutionally preprocessed using a cyclic convolution network (e.g., long and short memory networks LSTM, gated loop units GRU). Thus, long-term dependency and local patterns in the time sequence can be captured, and the input and output layer structures of the convolutional network can be determined, including defining time steps, feature dimensions and the like. And (3) properly selecting according to the complexity of the problem and the size of the data set, and extracting the characteristics and the labels from the demand trend prediction convolution sample set. The characteristics represent input data, the labels represent corresponding demand trend prediction results, and a proper data cutting mode is selected according to the demands of the problems and the model design. For example, the data sequence may be cut into non-overlapping time periods in a fixed time step, or a sliding window cut may be performed to obtain more samples, the cut data sequence may be used as an input to a convolutional network, and the corresponding demand trend prediction result may be used as a label. And ensuring that the corresponding relation between the data sequence and the label is correct, carrying out convolution operation on the demand trend prediction convolution sequence by using a cavity convolution algorithm (for example, a cavity convolution layer and expansion convolution), and setting proper cavity rate and convolution kernel size so as to adapt to feature extraction requirements on different time scales. The larger void ratio can increase receptive field, capture longer time dependence, pay attention to the depth and width of the convolution layer to meet model complexity and actual demands, and perform spatial pyramid pooling operation on the feature map extracted by the demand trend prediction convolution network. The spatial pyramid pooling can extract features on multiple scales, increase the adaptability of a model to variable scales, set proper pyramid levels, perform pooling operation on each level in scale and stride, obtain a multi-layer feature map after pooling operation, represent abstract features on different scales, and input a demand trend prediction convolution feature map into a combined classifier algorithm. The combined classifier can be an integrated learning method, such as a random forest, a gradient lifting tree and the like, and a plurality of classifiers are stacked and combined by considering a stacking integration method so as to improve the prediction performance and generalization capability, and the models are evaluated and selected through a cross-validation method and the like. And (3) performing model performance verification by using a test set, and ensuring that the model has better performance on unseen data.

The present embodiments are, therefore, to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein.

It will be understood that, although the terms "first," "second," etc. may be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another element. For example, a first element could be termed a second element, and, similarly, a second element could be termed a first element, without departing from the scope of example embodiments. The term "and/or" as used herein includes any and all combinations of one or more of the associated listed items.

The foregoing is only a specific embodiment of the invention to enable those skilled in the art to understand or practice the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. The medical surgical robot market current situation demand analysis method is characterized by comprising the following steps of:

Step S5: performing data visualization on the demand trend prediction data of the surgical robot to generate a demand trend prediction interactive view;

2. The method according to claim 1, wherein the specific steps of step S1 are:

3. The method according to claim 2, wherein the specific steps of step S2 are:

4. A method according to claim 3, wherein the specific step of step S3 is:

step S34: carrying out association rule mining on the demand feature cluster data to generate a demand feature association set;

5. The method according to claim 4, wherein the specific steps of step S32 are:

6. The method according to claim 5, wherein the specific step of step S4 is:

the time-series demand weight calculation formula in step S42 specifically includes:

7. The method according to claim 6, wherein the market demand trend prediction calculation formula in step S45 is specifically:

8. The method according to claim 7, wherein the specific steps of step S5 are:

9. The method according to claim 8, wherein the specific step of step S6 is:

10. A medical surgical robot market-state demand analysis system for performing the medical surgical robot market-state demand analysis method of claim 1, comprising:

the data visualization module is used for carrying out data visualization on the demand trend prediction data of the surgical robot by using a deep learning algorithm to generate a demand trend prediction interactive view;