Science and technology project risk monitoring method based on artificial intelligence

Abstract

The invention discloses a science and technology project risk monitoring method based on artificial intelligence, and relates to the technical field of project risk monitoring. The method comprises the steps of obtaining participants of a target science and technology project, collecting project contents corresponding to the participants, and setting one participant and the corresponding project contents as one project point to obtain a plurality of project points; obtaining the monitoring difficulty of the project points, and selecting the monitored project points as monitoring points; monitoring data of the monitoring points are collected, risk fluctuation values of the monitoring points are obtained through analysis, and monitoring time of the monitoring points is determined; forming a monitoring scheme for risk monitoring of the target science and technology project according to the monitoring points and the corresponding monitoring time, and executing the monitoring scheme; and obtaining a real-time monitoring risk of the monitoring point, performing analysis to obtain a total project risk of the target science and technology project, and sending the total project risk to the user side. According to the invention, the cost of science and technology project risk monitoring based on artificial intelligence is reduced.

Description

Technical Field

[0001] This application relates to the technical field of project risk monitoring, and in particular, to a risk monitoring method for scientific and technological projects based on artificial intelligence. Background Art

[0002] Risk monitoring of scientific and technological projects refers to the process of continuously identifying, evaluating, and responding to potential risks through a systematic approach during the implementation of scientific and technological projects, in order to ensure that the project progresses as planned and achieves its goals. Its core lies in dynamically tracking risk factors in key areas such as technology R & D, market changes, and resource allocation, and reducing the impact of risks on the project through early warning mechanisms and response strategies. With the development of technology, scientific and technological projects are naturally emerging in an endless stream. Therefore, risk monitoring of scientific and technological projects is conducive to timely discovering problems, thereby solving problems and promoting the development of scientific and technological projects. In the current existing risk monitoring of scientific and technological projects, risk assessment is achieved by collecting all the data of the project. By summarizing all the data of the project for scientific and technological project risk assessment, on the one hand, a large amount of monitoring resources are required, and on the other hand, processing and monitoring all the data will also be very cumbersome, increasing the complexity of data processing. Summary of the Invention

[0003] The purpose of the present invention is to provide a risk monitoring method for scientific and technological projects based on artificial intelligence to solve the problems raised in the above background art.

[0004] A risk monitoring method for scientific and technological projects based on artificial intelligence provided by this application adopts the following technical solutions: Obtain the participants of the target scientific and technological project, collect the project content corresponding to the participants, and set one participant and the corresponding project content as a project point to obtain multiple project points; Obtain the monitoring difficulty of the project points, and select the project points to be monitored as monitoring points according to the monitoring difficulty; Collect the monitoring data of the monitoring points, analyze the risk fluctuation value of the monitoring points based on the monitoring data, and determine the monitoring time of the monitoring points according to the risk fluctuation value; Form a monitoring plan for the risk monitoring of the target scientific and technological project based on the monitoring points and the corresponding monitoring time, and execute the monitoring plan; Obtain the real-time monitoring risk of the monitoring points, analyze the total project risk of the target scientific and technological project based on the real-time monitoring risk, and send it to the user terminal.

[0005] Preferably, the step of obtaining the monitoring difficulty of the project points and selecting the project points to be monitored as monitoring points according to the monitoring difficulty is specifically: Collect the project content of the project points, and judge the importance degree of the project points according to the project content; Collect the association relationship between the collection project point and other project points, and evaluate the independence degree of the project point according to the association relationship; Obtain the requirement data of the monitored project point, and evaluate the collection difficulty of the requirement data; Obtain the evaluation requirement information of the monitored project point, and judge the evaluation difficulty according to the evaluation requirement information; Analyze the monitoring difficulty of the project point by synthesizing the collection difficulty and the evaluation difficulty, and judge whether to monitor the project point by combining the independence degree and the importance degree; If the project point is to be monitored, then mark the project point as a monitored point.

[0006] Preferably, the steps of collecting the association relationship between the collection project point and other project points and evaluating the independence degree of the project point according to the association relationship are specifically as follows: Obtain the personnel association relationship between the personnel corresponding to the project point and the personnel corresponding to other project points, and obtain the personnel independence degree according to the personnel association relationship; Judge whether there is an association relationship between the project content of the project point and the project content of other project points. If there is an association relationship, then judge whether the risk change of the project point causes the risk change of other project points; If it causes the risk change of other project points, then count the risk changes of other project points, analyze to obtain the project independence degree, and combine the personnel independence degree to obtain the independence degree of the project point; If it does not cause the risk change of other project points, then the project independence degree is the maximum value.

[0007] Preferably, the steps of obtaining the personnel association relationship between the personnel corresponding to the project point and the personnel corresponding to other project points and obtaining the personnel independence degree according to the personnel association relationship are specifically as follows: Mark the personnel corresponding to the project point as the target personnel, and mark the personnel corresponding to other project points as other personnel; Count the number of other personnel associated with the target personnel and record it as the number of associated personnel; Count the communication frequency between the target personnel and other personnel, and count the proportion of shared information between the target personnel and other personnel; Combine the number of associated personnel, the communication frequency and the proportion of shared information to comprehensively obtain the personnel independence degree.

[0008] Preferably, the steps of if it causes the risk change of other project points, then count the risk changes of other project points and analyze to obtain the project independence degree are specifically as follows: Collect the risk change value of the project point, obtain the risk change value of other project points corresponding to the risk change value of the project point, and form a project risk change value - other risk change value table; Compare the other risk change values in the project risk change value - other risk change value table with the project risk change values to obtain the risk reflection values of other project points reflecting the risk changes of the project points; Statistically analyze the risk delay duration of the risk changes of other project points caused by the risk changes of the project points; Based on the risk reflection values and the risk delay duration, comprehensively obtain the risk visibility values of other project points, and sum the reciprocals of the risk visibility values of all other project points to obtain the project independence degree.

[0009] Preferably, the steps of obtaining the requirement data of the monitored project points and evaluating the acquisition difficulty of the requirement data are specifically as follows: Obtain the requirement data types of the monitored project points and determine whether the data corresponding to the requirement data types is recorded; If the corresponding data is recorded, then statistically analyze the required quantity of the recorded data, the number of steps for recording the data, and the number of personnel for recording the data, and comprehensively obtain the data recording difficulty; Collect the number of retrieval steps and the average retrieval duration of the retrieval requirement data types, and combine the number of retrieval steps and the average retrieval duration to obtain the data retrieval difficulty; Combine the data recording difficulty and the data retrieval difficulty to comprehensively obtain the type acquisition difficulty of the requirement data types; Statistically analyze the average requirement data volume of the requirement data types, combine the type acquisition difficulty to obtain the type acquisition difficulty of the requirement data types, and sum the type acquisition difficulties of all requirement data types to obtain the acquisition difficulty of the requirement data.

[0010] Preferably, the steps of obtaining the evaluation requirement information of the monitored project points and judging the evaluation difficulty according to the evaluation requirement information are specifically as follows: Obtain the evaluation requirement information of the monitored project points, and extract the subjective evaluation content according to the evaluation requirement information; Statistically analyze the ratio of the subjective evaluation content to all risk evaluation contents to obtain the subjective evaluation ratio; Extract the number of evaluation factors according to the subjective evaluation content, and combine the subjective evaluation ratio to obtain the content evaluation difficulty; Obtain the evaluators of the subjective evaluation content, collect the evaluation data of the evaluators, and obtain the evaluation accuracy of the evaluators according to the evaluation data; Combine the evaluation accuracy and the content evaluation difficulty to confirm the evaluation difficulty of the project points.

[0011] Preferably, the steps of obtaining the evaluators of the subjective evaluation content, collecting the evaluation data of the evaluators, and obtaining the evaluation accuracy of the evaluators according to the evaluation data are specifically as follows: Judge whether there are multiple evaluators. If there are multiple evaluators, then statistically analyze the number of evaluators; Statistically analyze the knowledge coverage of multiple evaluators, and confirm the evaluation accuracy in combination with the number of evaluators. If there are not multiple evaluators, obtain the historical evaluation accuracy rate of the evaluator as the evaluation accuracy.

[0012] Preferably, the steps of collecting the monitoring data of the monitoring point, analyzing the risk fluctuation value of the monitoring point according to the monitoring data, and determining the monitoring time of the monitoring point according to the risk fluctuation value are specifically as follows: Collect the monitoring data of the monitoring point, evaluate the risk of the project point according to the monitoring data, and form a corresponding risk fluctuation curve according to the risk of the project point; Calculate the risk difference between different time points according to the risk fluctuation curve and record it as the risk fluctuation value; Screen the average duration between time points where the risk fluctuation value is greater than the preset fluctuation threshold as the monitoring duration; Obtain other monitoring points where the risk visibility value reaches the preset risk visibility threshold as the visible monitoring points; Select the time point of the most recent monitoring without warning of all visible monitoring points as the most recent visible time point; Collect the time point of the most recent monitoring of the monitoring point, compare it with the most recent visible time point, and select the one closest to the real-time time point as the standard time point; According to the standard time point and the monitoring duration, confirm the next monitoring time point of the monitoring point.

[0013] Preferably, the steps of obtaining the real-time monitoring risk of the monitoring point, analyzing the total project risk of the target science and technology project according to the real-time monitoring risk, and sending it to the user terminal are specifically as follows: Set the monitoring risk threshold of the monitoring point. When the real-time monitoring risk reaches the monitoring risk threshold, send the monitoring data of the monitoring point to the user terminal; If the real-time monitoring risks of all monitoring points do not reach the monitoring risk threshold, evaluate the total project risk of the science and technology project according to the real-time monitoring risks of all monitoring points; Judge whether the total project risk reaches the monitoring risk threshold. If it reaches the monitoring risk threshold, send the total project risk to the user terminal and package the real-time monitoring risks of all monitoring points and send them to the user terminal.

[0014] In summary, the present application includes at least one of the following beneficial technical effects: 1. Set each participant and the corresponding project content as a project point, select the monitoring points for the risk monitoring of scientific and technological projects according to the monitoring difficulty of the project points, and then set the monitoring time according to the risk fluctuation of the monitoring points to form a monitoring plan. Monitor the scientific and technological projects according to the monitoring plan, and give early warnings and send data in a timely manner. By screening and arranging the monitoring points, on the one hand, distributed monitoring is realized, making the risk monitoring of scientific and technological projects more comprehensive. On the other hand, unnecessary risk monitoring resources are reduced, and the cost of risk monitoring of scientific and technological projects based on artificial intelligence is reduced.

[0015] 2. Collect the project content of the project points, and judge the importance of the project points according to the project content. According to the correlation between the participants, project content in the project points and other project points, evaluate the independence of the project points according to the correlation. Then, combined with the data collection difficulty and risk assessment difficulty of the project points, the monitoring points are screened. First, screening according to the importance reduces the impact of missing the monitoring of important content on the risk monitoring of scientific and technological projects, and the screening basis of the independence degree can reduce missed monitoring and improve the comprehensiveness of monitoring. The monitoring difficulty can reduce unnecessary resource waste and improve the comprehensiveness and accuracy of the risk monitoring of scientific and technological projects based on artificial intelligence.

[0016] 3. For the measurement of the evaluation difficulty, both the proportion of subjective content in the evaluation and the ability of the evaluator are considered. It not only reduces the interference of subjective evaluation on the evaluation, but also reduces the risk of incorrect evaluation results caused by the insufficient ability of the evaluator. Among the monitoring points screened by the evaluation difficulty, the interference of invalid and incorrect evaluation results on the results of the risk monitoring of scientific and technological projects is greatly reduced, and the accuracy of the risk monitoring of scientific and technological projects based on artificial intelligence is improved. Description of the Drawings

[0017] Figure 1 It is a schematic diagram of the specific steps of an embodiment of a method for risk monitoring of scientific and technological projects based on artificial intelligence according to the present invention. Detailed Embodiment

[0018] The following combines the embodiments and Figure 1 Further detailed description of the present invention is given, but the implementation manners of the present invention are not limited thereto.

[0019] The present invention discloses a method for risk monitoring of scientific and technological projects based on artificial intelligence, which specifically includes the following steps: Step S1, obtain the participants of the target scientific and technological project, collect the project content corresponding to the participants, set one participant and the corresponding project content as a project point, and obtain multiple project points.

[0020] Science and technology projects usually cover multiple sectors and multiple personnel. Different personnel are responsible for different contents of science and technology projects, forming a complete science and technology project. Setting each participating personnel and the corresponding project content as a project point, multiple project points are obtained.

[0021] Step S2: Obtain the monitoring difficulty of the project points, and select the project points to be monitored according to the monitoring difficulty as the monitoring points.

[0022] Step S3: Collect the monitoring data of the monitoring points, analyze the risk fluctuation value of the monitoring points based on the monitoring data, and determine the monitoring time of the monitoring points according to the risk fluctuation value.

[0023] Step S4: Form a monitoring plan for the risk monitoring of the target science and technology project based on the monitoring points and the corresponding monitoring time, and execute the monitoring plan.

[0024] Step S5: Obtain the real-time monitoring risk of the monitoring points, analyze the total project risk of the target science and technology project based on the real-time monitoring risk, and send it to the user terminal.

[0025] In practical applications, for the risk monitoring of traditional science and technology projects, it is necessary to collect all data and then conduct risk analysis to achieve monitoring. However, due to the large scale of science and technology projects, involving a large number of personnel and a large amount of content, and some monitoring data has little effect on this risk monitoring. Therefore, comprehensive monitoring increases the monitoring workload on the one hand, thus increasing the waste of monitoring resources. On the other hand, some ineffective and incorrect monitoring contents are likely to lead to errors and inaccuracies in the risk monitoring effect. By selecting monitoring points for monitoring, not only distributed monitoring is achieved, but also some ineffective, incorrect, and less useful monitoring data is reduced, greatly saving the cost and resources of the risk monitoring of science and technology projects.

[0026] The step of obtaining the monitoring difficulty of the project points and selecting the project points to be monitored according to the monitoring difficulty as the monitoring points is specifically as follows: Step S21: Collect the project content of the project points, and judge the importance degree of the project points according to the project content.

[0027] The importance degree of the project can be obtained through user input, or the scores of the project points can be comprehensively calculated by setting multiple evaluation dimensions and assigning weights.

[0028] Step S22: Collect the association relationship between the project points and other project points, and evaluate the independence degree of the project points according to the association relationship.

[0029] Step S23: Obtain the requirement data for monitoring the project points, and evaluate the collection difficulty of the requirement data.

[0030] Step S24: Obtain the evaluation requirement information of the monitored project point, and judge the evaluation difficulty based on the evaluation requirement information.

[0031] Step S25: Analyze the collection difficulty and the evaluation difficulty comprehensively to obtain the monitoring difficulty of the project point, and judge whether to monitor the project point in combination with the independence degree and the importance degree.

[0032] Step S26: If the project point is to be monitored, record the project point as a monitored point.

[0033] In practical applications, the weight ratios of the collection difficulty and the evaluation difficulty are set respectively, and the monitoring difficulty is calculated according to the weight ratios. For example, if the weight ratios of the collection difficulty and the evaluation difficulty are set as 40% and 60% respectively, and the collection difficulty and the evaluation difficulty are 20 and 50 respectively, then the monitoring difficulty is 20×40% + 50×60% = 38. Use the analytic hierarchy process to evaluate whether to monitor the project point. By constructing a hierarchical structure model (goal layer, criterion layer, index layer), combine expert experience to judge the relative importance of the monitoring difficulty, the independence degree and the importance degree, calculate the global weight and the priority ranking, and obtain the monitoring value of the project point. If the monitoring value reaches the preset monitoring value threshold, then monitor this point and record it as a monitored point. The larger the monitoring value, the more this point needs to be monitored. Among them, the greater the monitoring difficulty, the smaller the monitoring value, because the greater the monitoring difficulty, the more resources are consumed, and the accuracy of the monitored data may also decrease. Therefore, it is less suitable as a monitored point. The greater the independence degree, the more it needs to be monitored, because the risk of this point cannot be confirmed by other monitoring data, so this point needs to be arranged as a monitored point. And the greater the importance degree, the more it needs to be monitored, and the corresponding monitoring value is larger, because it is related to the risk of the entire scientific and technological project.

[0034] The steps of collecting the correlation relationship between the project point and other project points and evaluating the independence degree of the project point according to the correlation relationship are specifically as follows: Step S221: Obtain the personnel correlation relationship between the personnel corresponding to the project point and the personnel corresponding to other project points, and obtain the personnel independence degree according to the personnel correlation relationship.

[0035] Step S222: Judge whether there is a correlation relationship between the project content of the project point and the project content of other project points. If there is a correlation relationship, judge whether the risk change of the project point causes the risk change of other project points.

[0036] Step S223: If it causes the risk change of other project points, then count the risk change of other project points, analyze to obtain the project independence degree, and combine the personnel independence degree to obtain the independence degree of the project point.

[0037] Set the weight ratios of the project independence degree and the personnel independence degree respectively, and calculate the independence degree of the project point according to the weight ratios. For example, set the weight ratios of the project independence degree and the personnel independence degree to 80% and 20% respectively, and the project independence degree and the personnel independence degree are 40% and 20% respectively. Then the independence degree of the project point is 40%×80% + 20%×20% = 36%.

[0038] Step S224, if it does not cause the risk change of other project points, the project independence degree is the maximum value.

[0039] In practical applications, the degree of association between project points is considered from two aspects: on the one hand, the degree of association of the project point content, and on the other hand, the degree of association between the participants of the project points. If there is no association between the project point contents, it is defaulted that the project independence degree is the maximum value. For example, for a technology project of a massager, Project Point A makes a heating module and Project Point B makes a voice module. There is no association between the two module contents, and both are only associated with the main control system content. Then the project independence degree between Project Points A and B is 100%, which belongs to complete independence. If there is an association relationship between two project points, it means that the project content may not be completely independent. According to the fluctuation of historical risk data, it can be found whether the risk of the remaining project points changes after the risk of each project point changes, so as to confirm whether the project point causes the risk change of other project points. If the project point does not cause the risk change of other project points, it means that although there is an association between the project point contents, the risk of the project point cannot be monitored through other project points. Therefore, it is also considered that the project point is completely independent in risk monitoring, so the project independence degree is the maximum value. For example, Project Point A makes a heating module and Project Point B makes a control module. There is content interaction between the two, and the heating needs to be controlled. However, when the heating module has a risk, it does not affect the control function of the control module, and the risk of the control module does not change. So the heating module does not cause the risk change of the control module, and it is impossible to infer whether the heating module has a risk through the risk change of the control module. Therefore, the project independence degree is the maximum value.

[0040] The steps to obtain the personnel association relationship between the participants corresponding to the project point and the participants corresponding to other project points and obtain the personnel independence degree according to the personnel association relationship are as follows: Step S2211, record the participants corresponding to the project point as target personnel, and record the participants corresponding to other project points as other personnel.

[0041] Step S2212, count the number of other personnel related to the target personnel and record it as the number of associated personnel.

[0042] Step S2213, count the communication frequency between the target personnel and other personnel, and count the proportion of shared information between the target personnel and other personnel.

[0043] The proportion of shared information refers to the ratio of the content of shared information between the target person and other persons to the total information content of the target person.

[0044] Step S2214, combining the number of associated persons, communication frequency, and proportion of shared information, comprehensively obtain the degree of personnel independence.

[0045] In practical applications, there may be no association between the contents of some project points, but there is associated management between the personnel, which will also affect the degree of independence of the project points. When the degree of personnel independence is low, it reflects that communication is needed between two project points to promote the development of the project through communication. Then, the degree of independence between the project points will be reduced at this time. The more the number of associated persons of the participants in a project point with the participants in other project points, the higher the degree of association with the entire project, and thus the lower the degree of personnel independence. Through the existing weight scoring model, according to the number of associated persons, communication frequency, and proportion of shared information, the degree of personnel independence is obtained. Among them, when the number of associated persons is larger, the communication frequency is larger, and the proportion of shared information is larger, the degree of personnel independence is lower. For example, Project Point A is responsible for the heating module in the technology project product. Its participants need to communicate with multiple persons such as salespersons, solution designers, and control module personnel, indicating that the heating module is relatively important and has a high degree of association with other project points, thereby reducing the degree of independence of the project point. If Project Point A does not need to communicate with other persons, it indicates that the heating module may be common existing technology in the market, so no communication is required and it does not affect the risks of other project points. Therefore, the degree of personnel independence is high, and the degree of independence of the associated project point is even higher.

[0046] If it causes changes in the risks of other project points, then the steps of statistically analyzing the risk changes of other project points and obtaining the degree of project independence are as follows: Step S2231, collect the risk change values of the project points, obtain the risk change values of other project points corresponding to the risk change values of the project points, and form a project risk change value - other risk change value table.

[0047] Collect the risk change values of the project points at different time points, and statistically analyze the risk change values of other project points at the corresponding time points. For example, the risk change value of Project Point A is 30%, the risk change value of Project Point B is 60%, and the risk change value of Project Point C is 20%.

[0048] Step S2232, compare the other risk change values and the project risk change values in the project risk change value - other risk change value table to obtain the risk reflection value of other project points reflecting the risk changes of the project points.

[0049] For example, compare the risk change value of Project Point A, which is 30%, with the risk change value of Project Point B, which is 60%. We get 60% / 30% = 200%, so the risk reflection value is 200%. This indicates that Project Point B can reflect the risk change of Project Point A to a greater extent.

[0050] Step S2233: Statistically calculate the risk delay duration of the risk change of other project points caused by the risk change of a project point.

[0051] Not all risks will immediately cause risks to other project points. Some risks are delayed. Based on historical risk data, confirm the delay duration of the previous risk change of a project point. For example, when a risk change occurs at Project Point A and a risk change occurs at Project Point B on average one week later and it is confirmed that it is caused by the risk change of Project Point A, then the risk delay duration is one week.

[0052] Step S2234: Comprehensively obtain the risk visibility value of other project points based on the risk reflection value and the risk delay duration. Sum the reciprocals of the risk visibility values of all other project points to obtain the project independence degree.

[0053] In practical applications, set the weight ratios of the risk reflection value and the risk delay duration respectively, and calculate the risk visibility value of other project points according to the weight ratios. For example, set the weight ratios of the risk reflection value and the risk delay duration to 50% and 50% respectively. If the risk reflection value and the risk delay duration are 60% and 7 days respectively, then the risk visibility value is 60% × 50% + 7 × 50% = 3.8. Sum the reciprocals of the risk visibility values to obtain the result as the project independence degree. For example, if the risk visibility values are 4, 2, and 8, then the project independence degree is 1 / 4 + 1 / 2 + 1 / 8 = 0.875. When the risk of a project point can be monitored through the risks of other project points, then it is considered that the independence degree of this project point in risk monitoring is relatively low because the risk monitoring of this project point can be achieved through other project points. The lower the project independence degree, the lower the independence degree of the project point will be. Therefore, the monitoring requirement will be lower because the risk change of this project point can also be inferred through other project points, so there is no need for additional monitoring.

[0054] The steps to obtain the requirement data of the monitored project point and evaluate the acquisition difficulty of the requirement data are specifically as follows: Step S231: Obtain the type of requirement data of the monitored project point and determine whether the data corresponding to the requirement data type is recorded.

[0055] During the process of monitoring project points, not all required data will be recorded. For example, when monitoring Project Point A, it is necessary to know the research records of the duration researchers. However, the research records are obtained through conversation research, so the corresponding conversation research data of this required data type is not recorded. If it is necessary to develop error reporting data, the corresponding data of this required data type has been recorded.

[0056] Step S232, if the corresponding data is recorded, count the required quantity of the recorded data, the number of steps of the recorded data, and the number of personnel of the recorded data, and comprehensively obtain the data recording difficulty.

[0057] Respectively set the weight ratios of the required quantity of the recorded data, the number of steps of the recorded data, and the number of personnel of the recorded data, and calculate the data recording difficulty according to the weight ratios. If the data is not recorded, it is defaulted that the data recording difficulty is the maximum value. When the required quantity is larger, the number of steps is larger, and the number of personnel is larger, it indicates that more resources are required to record the data, and the difficulty of collecting the data increases.

[0058] Step S233, collect the number of retrieval steps and the average retrieval duration of the retrieved required data type, and combine the number of retrieval steps and the average retrieval duration to obtain the data retrieval difficulty.

[0059] Respectively set the proportionality coefficients of the number of retrieval steps and the average retrieval duration, multiply the number of retrieval steps and the average retrieval duration by the corresponding proportionality coefficients and then sum them up to obtain the data retrieval difficulty. Some data is easy to collect, but due to issues such as technical secrets and user privacy, the difficulty of retrieving this data is relatively large. Then, for risk monitoring, it will increase its monitoring difficulty. For example, some data retrieval requires multiple levels of approval from senior leaders. Then, for risk monitoring, a large amount of time is required to retrieve the data, increasing the retrieval difficulty.

[0060] Step S234, comprehensively obtain the type acquisition difficulty of the required data type by combining the data recording difficulty and the data retrieval difficulty.

[0061] Determine the weights of the data recording difficulty and the retrieval difficulty through expert scoring, and finally calculate the comprehensive acquisition difficulty value by weighted calculation.

[0062] Step S235, count the average required data volume of the required data type, combine the type acquisition difficulty to obtain the type collection difficulty of the required data type, sum up the type collection difficulties of all required data types, and obtain the collection difficulty of the required data.

[0063] In actual application, a linear equation based on historical data is constructed for the collection difficulty of types, the average demand data volume, and the acquisition difficulty of types. The coefficients are fitted by the least squares method to quantify the contribution weights of the average demand data volume and the acquisition difficulty of types to the collection difficulty of types. The collection difficulty of types is calculated based on the constructed linear equation, and then the collection difficulties of all demand data types are summed up to obtain the collection difficulty of the data. When the data collection difficulty is greater and the average demand data volume is larger, the collection difficulty of this type will increase significantly. Since in the risk monitoring of scientific and technological projects, multiple aspects of data are usually required for monitoring and analysis, there are also multiple types of required data. The collection difficulties of all demand data types required to monitor this project point are superimposed to confirm the collection difficulty of the demand data.

[0064] The steps to obtain the evaluation requirement information of the monitored project point and judge the evaluation difficulty according to the evaluation requirement information are as follows: Step S241: Obtain the evaluation requirement information of the monitored project point and extract the subjective evaluation content according to the evaluation requirement information.

[0065] In some evaluations, personnel intervention is required or it depends on existing personnel experience. Then, according to the evaluation requirement information, the evaluation content related to personnel is evaluated as subjective content.

[0066] Step S242: Calculate the ratio of the subjective evaluation content to all risk evaluation contents to obtain the subjective evaluation ratio.

[0067] Step S243: Extract the number of evaluation factors according to the subjective evaluation content, and combine it with the subjective evaluation ratio to obtain the content evaluation difficulty.

[0068] The weight ratios of the number of evaluation factors and the subjective evaluation ratio are set respectively, and the content evaluation difficulty is calculated according to the weight ratios. When the proportion of subjective content is larger, it is more difficult for the machine equipment to conduct risk assessments of scientific and technological projects from a rational and objective perspective, so the evaluation difficulty is greater. And when the number of evaluation factors is larger, it means that more directions and aspects are considered, and the difficulty for the machine equipment to give risk monitoring results through big data analysis will also increase.

[0069] Step S244: Obtain the evaluators of the subjective evaluation content, collect the evaluation data of the evaluators, and obtain the evaluation accuracy of the evaluators according to the evaluation data.

[0070] Step S245: Combine the evaluation accuracy and the content evaluation difficulty to confirm the evaluation difficulty of the project point.

[0071] In practical applications, the weight ratios of the evaluation accuracy and the content evaluation difficulty are set respectively, and the evaluation difficulty of the project points is calculated according to the weight ratios. When the evaluation accuracy is higher, the evaluation difficulty of the project points is smaller, while when the content evaluation difficulty is greater, the evaluation difficulty of the project points is also greater. On the one hand, the more subjective content the content involves, the greater the probability of errors in the objectively given risk assessment, thus increasing the evaluation difficulty of the system. For the personnel evaluation involved, when collecting the evaluation results of personnel for risk monitoring and analysis, it is necessary to consider whether the personnel evaluation is accurate. If the personnel evaluation is inaccurate, it will obviously directly affect the final risk monitoring result. If the personnel evaluation is inaccurate, for the system, it is more difficult to process the error information for monitoring and analysis, and the risk assessment is more difficult.

[0072] The steps to obtain the evaluators of the subjective evaluation content, collect the evaluation data of the evaluators, and obtain the evaluation accuracy of the evaluators according to the evaluation data are specifically as follows: Step S2441: Determine whether there are multiple evaluators. If there are multiple evaluators, count the number of evaluators.

[0073] Step S2442: Statistically analyze the knowledge coverage of multiple evaluators, and confirm the evaluation accuracy in combination with the number of evaluators.

[0074] Step S2443: If there are no multiple evaluators, obtain the historical evaluation correct rate of the evaluator as the evaluation accuracy.

[0075] In practical applications, the weight ratio values of the knowledge coverage and the number of evaluators are set respectively, and the evaluation accuracy is calculated according to the weight ratio values. When the knowledge coverage is wider, it means that the evaluators consider more aspects, and the obtained evaluation accuracy will increase. The more evaluators there are, the easier it is to pool wisdom, identify omissions and make up for deficiencies, and reduce the occurrence of evaluation errors. Therefore, the evaluation accuracy of the evaluators is higher. When there are no multiple evaluators and only one evaluator, the historical evaluation correct rate of this evaluator is directly used as the evaluation accuracy.

[0076] The steps to collect the monitoring data of the monitoring points, obtain the risk fluctuation value of the monitoring points according to the monitoring data analysis, and determine the monitoring time of the monitoring points according to the risk fluctuation value are specifically as follows: Step S31: Collect the monitoring data of the monitoring points, evaluate the project point risks according to the monitoring data, and form a corresponding risk fluctuation curve based on the project point risks.

[0077] Step S32: Calculate the risk difference between different time points according to the risk fluctuation curve and record it as the risk fluctuation value.

[0078] Step S33: Screen the average duration between the time points with the risk fluctuation value greater than the preset fluctuation threshold as the monitoring duration.

[0079] Step S34: Obtain other monitoring points whose risk visibility values reach the preset risk visibility value threshold, which are the visible monitoring points.

[0080] Step S35: Select the time point of the most recent monitoring without warning for all visible monitoring points and record it as the most recent visible time point.

[0081] For example, if the visibility value of monitoring point A to monitoring point B is 200%, and monitoring point A is monitored at 10 o'clock without any monitoring warning, it indicates that monitoring point B is also normal. Here, the most recent time is relative to the monitoring point and is calculated based on the risk extension duration. That is to say, if monitoring point A is monitored at 10 o'clock and the risk extension duration is 1 hour, then the time point of the most recent monitoring without warning is considered to be 9 o'clock because for the monitoring point, the risk monitored at 10 o'clock is actually the risk at 9 o'clock.

[0082] Step S36: Collect the time point of the most recent monitoring of the monitoring point, compare it with the most recent visible time point, and select the one closest to the real-time time point as the standard time point.

[0083] Step S37: Confirm the next monitoring time point of the monitoring point according to the standard time point and the monitoring duration.

[0084] In practical applications, the risks of monitoring project points do not need to be monitored in real time. For some monitoring points with little risk fluctuation and no special risks throughout the year, the monitoring can be carried out at a longer interval. For some project points with large risk fluctuations, close monitoring is required to reduce the impact caused by missed monitoring. The average duration when the risk fluctuation difference reaches the threshold is used as the monitoring duration. As mentioned in the independence of the previous project points, the risks of some project points can be inferred from the monitoring risks of other project points. Then, if no abnormalities occur in the monitoring of other project points, the monitoring can be postponed. For example, the time point of the most recent monitoring of the monitoring point is 10 o'clock, and the monitoring duration is 1 hour per time. At 10:30, monitoring data of monitoring point A is collected without any abnormalities, and monitoring point B is monitored at 10:40 without any risk warning. Then, the closest time point is located at 10:40, and the monitoring point can be monitored again at 11:40 for the next time.

[0085] The steps of obtaining the real-time monitoring risk of the monitoring point, analyzing the total project risk of the target science and technology project based on the real-time monitoring risk, and sending it to the user terminal are specifically as follows: Step S51: Set the monitoring risk threshold of the monitoring point. When the real-time monitoring risk reaches the monitoring risk threshold, send the monitoring data of the monitoring point to the user terminal.

[0086] Step S52: If the real-time monitoring risks of all monitoring points do not reach the monitoring risk threshold, evaluate the total project risk of the science and technology project based on the real-time monitoring risks of all monitoring points.

[0087] The overall project risk of a scientific and technological project can be obtained through expert evaluation rules combined with the real-time monitoring risk assessment of all monitoring points. Alternatively, the weight ratios of the real-time monitoring risks of different monitoring points can be set, and the overall project risk can be calculated based on the weight ratios.

[0088] In step S53, it is determined whether the overall project risk reaches the monitoring risk threshold. If it reaches the monitoring risk threshold, the overall project risk is sent to the user terminal, and the real-time monitoring risks of all monitoring points are packaged and sent to the user terminal.

[0089] In practical applications, through the selection of monitoring points for monitoring, distributed monitoring is achieved. This can not only timely detect the risk problems of each module but also confirm the overall project risk. When the risk of a project point is too high, it is necessary to timely adjust the content of that project point to reduce the impact on the overall project. When the risk of no monitoring point exceeds the threshold, the overall project risk is evaluated to confirm whether the overall risk of the scientific and technological project reaches the threshold. By monitoring each project point while monitoring the overall project with relatively small resources, the cost of scientific and technological project risk monitoring is saved.

[0090] The above are all the preferred embodiments of this application, and the protection scope of this application is not limited accordingly. Therefore, all equivalent changes made according to the structure, shape, and principle of this application should be covered within the protection scope of this application.

Claims

1. A scientific and technological project risk monitoring method based on artificial intelligence, characterized in that: The following steps are involved: Obtain participants of the target scientific and technological project, collect project contents corresponding to the participants, set one participant and the corresponding project content as one project point, and obtain multiple project points; Obtain the monitoring difficulty of the project point, and select the project point to be monitored as the monitoring point according to the monitoring difficulty; Collect monitoring data of monitoring points, obtain risk fluctuation values ​​of monitoring points according to monitoring data analysis, and determine monitoring time of monitoring points according to risk fluctuation values; Form a monitoring plan for risk monitoring of target science and technology projects based on monitoring points and corresponding monitoring time, and implement the monitoring plan; Obtain the real-time monitoring risk of the monitoring point, obtain the total project risk of the target technology project based on the real-time monitoring risk analysis, and send it to the user end.

2. According to the artificial intelligence-based scientific and technological project risk monitoring method of claim 1, it is characterized in that: The step of obtaining the monitoring difficulty of the project point and selecting the project point to be monitored as the monitoring point according to the monitoring difficulty is specifically: Collect the project content of the project point and determine the importance of the project point based on the project content; Collect the correlation between the project point and other project points, and evaluate the independence of the project point based on the correlation; Obtain demand data for monitoring project points and assess the difficulty of collecting demand data; Obtain the assessment requirement information of the monitoring project points, and determine the assessment difficulty based on the assessment requirement information; The monitoring difficulty of the project site is obtained by analyzing the difficulty of comprehensive data collection and evaluation, and whether the project site should be monitored is determined based on the degree of independence and importance; If the project point is monitored, the project point will be recorded as the monitoring point.

3. According to the artificial intelligence-based scientific and technological project risk monitoring method of claim 2, it is characterized in that: The step of collecting the correlation between the project point and other project points and evaluating the independence of the project points according to the correlation is specifically as follows: Obtain the personnel association relationship between the participants corresponding to the project point and the participants corresponding to other project points, and obtain the degree of personnel independence based on the personnel association relationship; Determine whether the project content of the project site is related to the project content of other project sites. If so, determine whether the risk changes of the project site will lead to risk changes of other project sites. If it causes risk changes in other project sites, the risk changes in other project sites will be counted, and the project independence degree will be obtained by analysis, and the independence degree of the project site will be obtained by combining the independence degree of the personnel; If it does not lead to changes in the risks of other project points, the degree of project independence is maximum.

4. According to the artificial intelligence-based scientific and technological project risk monitoring method of claim 3, it is characterized in that: The step of obtaining the personnel association relationship between the participants corresponding to the project point and the participants corresponding to other project points, and obtaining the degree of personnel independence according to the personnel association relationship, is specifically as follows: The participants corresponding to the project points are recorded as target personnel, and the participants corresponding to other project points are recorded as other personnel; The number of other persons who are related to the target person is counted as the number of related persons; Count the communication frequency between the target personnel and other personnel, and count the proportion of shared information between the target personnel and other personnel; The degree of personnel independence is comprehensively obtained by combining the number of related personnel, communication frequency and the proportion of shared information.

5. According to the artificial intelligence-based scientific and technological project risk monitoring method of claim 3, it is characterized in that: If the above causes the risk changes of other project points, the steps of statistically analyzing the risk changes of other project points and obtaining the degree of project independence are as follows: Collect the risk change value of the project point, obtain the risk change values ​​of other project points corresponding to the risk change value of the project point, and form a table of project risk change value-other risk change values; Compare the other risk change values ​​in the table of project risk change value - other risk change values ​​with the project risk change value to obtain the risk reflection values ​​of other project points that reflect the risk changes of the project points; Count the risk delay duration of other project point risk changes caused by project point risk changes; The risk visible values ​​of other project points are obtained based on the risk reflection value and the risk delay time. The reciprocals of the risk visible values ​​of all other project points are summed up to obtain the project independence degree.

6. According to claim 2, a method for monitoring scientific and technological project risks based on artificial intelligence is characterized in that: The steps of obtaining the demand data of the monitoring project points and evaluating the difficulty of obtaining the demand data are specifically as follows: Obtain the required data type of the monitoring project point, and determine whether the data corresponding to the required data type is recorded; If the corresponding data is recorded, the number of requirements for recording data, the number of steps for recording data, and the number of people who record data are counted to obtain the difficulty of data recording; Collect the number of retrieval steps and average retrieval time for retrieving the required data type, and combine the number of retrieval steps and average retrieval time to obtain the data retrieval difficulty; The difficulty of obtaining the required data type is obtained by combining the difficulty of data recording and the difficulty of data retrieval; The average amount of required data of the required data type is counted, and the type collection difficulty of the required data type is obtained by combining the type acquisition difficulty. The type collection difficulties of all required data types are summed to obtain the collection difficulty of the required data.

7. According to claim 2, a method for monitoring scientific and technological project risks based on artificial intelligence is characterized in that: The step of obtaining the assessment requirement information of the monitoring project point and determining the assessment difficulty according to the assessment requirement information is specifically as follows: Obtain the assessment requirement information of the monitoring project points, and extract the subjective assessment content based on the assessment requirement information; Calculate the ratio of subjective assessment content to all risk assessment content to obtain the subjective assessment ratio; The number of evaluation factors is extracted based on the subjective evaluation content, and the difficulty of content evaluation is obtained by combining the subjective evaluation ratio; Obtaining the evaluator of the subjective evaluation content, collecting the evaluation data of the evaluator, and obtaining the evaluation accuracy of the evaluator based on the evaluation data; The assessment accuracy and content assessment difficulty are combined to determine the assessment difficulty of the project point.

8. According to the artificial intelligence-based scientific and technological project risk monitoring method of claim 7, it is characterized in that: The steps of obtaining the evaluator of the subjective evaluation content, collecting the evaluation data of the evaluator, and obtaining the evaluation accuracy of the evaluator according to the evaluation data are specifically as follows: Determine whether there are multiple evaluators. If there are multiple evaluators, count the number of evaluators. Count the knowledge coverage of multiple evaluators and confirm the evaluation accuracy by combining the number of evaluators; If there are not multiple evaluators, the historical evaluation accuracy of the evaluator is obtained as the evaluation accuracy.

9. According to the artificial intelligence-based risk monitoring method for scientific and technological projects in claim 5, it is characterized in that: The steps of collecting monitoring data of the monitoring point, obtaining the risk fluctuation value of the monitoring point according to the monitoring data analysis, and determining the monitoring time of the monitoring point according to the risk fluctuation value are specifically: Collect monitoring data from monitoring points, evaluate project point risks based on the monitoring data, and form corresponding risk fluctuation curves based on project point risks; The risk difference between different time points is calculated based on the risk fluctuation curve and recorded as the risk fluctuation value; The average duration between the time points when the risk fluctuation value is greater than the preset fluctuation threshold is used as the monitoring duration; Other monitoring points whose risk visibility values ​​reach the preset risk visibility value threshold are visual monitoring points; The time point of the most recent monitoring without warning for all visible monitoring points is selected as the most recent visible time point; Collect the most recent monitoring time point of the monitoring point, compare it with the most recent visible time point, and select the time point closest to the real-time time point as the standard time point; According to the standard time point and monitoring duration, the next monitoring time point of the monitoring point is confirmed.

10. The method for monitoring scientific and technological project risks based on artificial intelligence according to claim 1 is characterized in that: The step of obtaining the real-time monitoring risk of the monitoring point, obtaining the total project risk of the target scientific and technological project according to the real-time monitoring risk analysis, and sending it to the user end is specifically as follows: Set the monitoring risk threshold of the monitoring point. When the real-time monitoring risk reaches the monitoring risk threshold, the monitoring data of the monitoring point will be sent to the user end; If the real-time monitoring risk of all monitoring points does not reach the monitoring risk threshold, the total project risk of the science and technology project is obtained based on the real-time monitoring risk assessment of all monitoring points; Determine whether the total risk of the project reaches the monitoring risk threshold. If so, send the total risk of the project to the user end, and package the real-time monitoring risks of all monitoring points and send them to the user end.

Images