A real-time early warning analysis method for research and development path technical conflicts

Abstract

The application discloses a kind of real-time early warning analysis methods for research and development path technical conflict, it is related to research and development management and technical information analysis technical field.The present application includes obtaining the research and development path data of target research and development path, element extraction and identification uniformity are carried out to the research and development path data, and generate research and development path element set.The present application is by element extraction and identification uniformity to target research and development path data, constructs the research and development path relationship chain including dependency relationship, constraint relationship, substitution relationship, mutual exclusion relationship and verification undertaking relationship, and is combined with path matching benchmark to carry out positioning, influence domain tracking and conflict propagation path analysis to path change event, the real-time identification and early warning of technical node matching relationship destruction in research and development path are realized, can timely find the technical conflict caused by target, constraint, parameter, data source, method selection or verification result change in research and development process, improve the timeliness and accuracy of research and development path risk identification.

Description

Technical Field

[0001] This invention belongs to the field of R&D management and technical intelligence analysis technology, and in particular relates to a real-time early warning analysis method for R&D path technology conflicts. Background Technology

[0002] Existing R&D activities typically follow a development path centered around predetermined technical objectives. This path includes stages such as technology selection, task decomposition, data collection, model building, experimental verification, result evaluation, path adjustment, and outcome consolidation. The R&D process generates a large amount of process data, technical documents, experimental records, parameter information, version information, external materials, and review comments. In existing technologies, R&D management systems record task nodes, execution status, personnel assignments, and stage results; data management systems store R&D data, test data, and verification results; document management systems collect technical materials, solution documents, and review materials; and knowledge management systems organize the relationships between concepts, indicators, methods, components, algorithms, and technical objects. Some technical solutions also incorporate text analysis, semantic recognition, knowledge graphs, rule models, and risk indicator models to extract and organize technical elements from R&D materials, forming relationships between technical nodes, technical features, R&D stages, and external information. R&D personnel use existing systems to view the execution status of the R&D path, organize R&D process materials, track changes in technical elements, and determine the matching relationship between the R&D direction and technical objectives during stage reviews, solution demonstrations, and path debriefings.

[0003] The core deficiency of existing technologies lies in the lack of a real-time early warning mechanism for technical conflicts in the R&D path. Existing systems are mostly used for recording R&D data, managing task status, and conducting phased analysis. Technical objectives, constraints, parameters, data sources, methodologies, and verification results are scattered across different data objects, making it difficult for the system to establish a continuous relationship for identifying technical conflicts along the R&D path. When the R&D path changes, existing systems cannot promptly determine whether the changes have disrupted the original technical path's compatibility. R&D personnel typically only discover conflicts during review, verification failures, or path adjustments. This delayed problem discovery and high path correction costs make it difficult to meet the need for early identification and real-time early warning of technical conflicts during the R&D process. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a real-time early warning analysis method for technical conflicts in R&D paths. The technical problem this invention aims to solve is: how to address the issues of difficulty in timely identification, determination of the scope of impact, difficulty in quantifying the degree of conflict, and delayed early warning after changes in R&D paths by unifying R&D path elements, constructing relationship chains, monitoring changes, tracking conflicts, calculating conflict values, and updating early warnings.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a real-time early warning and analysis method for R&D path technology conflicts, comprising: S1. Obtain the R&D path data of the target R&D path, extract and unify the elements of the R&D path data, and generate an R&D path element set; the R&D path element set includes execution status elements and matching relationship elements; S2. Construct a research and development path relationship chain based on the research and development path element set, the research and development path relationship chain including node relationships used to characterize the matching relationship between technology nodes; map the research and development path elements in the research and development path element set to the corresponding technology nodes to generate a path matching benchmark; S3. Monitor path change events in the target R&D path, locate the path change events to the corresponding changed technology nodes, and determine the node influence domain and conflict propagation path of the changed technology nodes based on the R&D path relationship chain; S4. Compare the path change event, the node influence domain, and the conflict propagation path with the path matching benchmark according to the conflict determination rules corresponding to the node relationship to generate a node conflict factor; calculate the R&D path conflict value based on the node conflict factor, whereby the R&D path conflict value characterizes the degree of damage the path change event causes to the target R&D path matching relationship; S5. When the R&D path conflict value reaches the preset warning threshold, generate a technical conflict warning result, and update the path matching benchmark and the R&D path relationship chain according to the confirmed conflict handling result.

[0006] The present invention is further configured such that the execution status elements include a technology node identifier, a node execution status, a node version number, and a node timing marker; the matching relationship elements include a technology target item, a technology constraint item, a parameter condition item, a data source item, a method selection item, and a verification result item; the unified element extraction and identification includes: identifying the execution status elements and the matching relationship elements from the R&D path data, generating an element record for each R&D path element, and merging different names, different format parameters, and different version identifiers representing the same technology object under the same standard identifier.

[0007] The present invention is further configured such that the node relationships include dependency relationships, constraint relationships, substitution relationships, mutual exclusion relationships, and verification succession relationships; the R&D path relationship chain is a directed relationship graph with technical nodes as node objects and the node relationships as edge objects; the edge objects include a starting technical node identifier, an ending technical node identifier, a node relationship type, a relationship direction, and a relationship constraint expression; the relationship constraint expression is used to record the matching conditions between two technical nodes on technical target items, technical constraint items, parameter condition items, data source items, method selection items, or verification result items.

[0008] The present invention is further configured such that the generation of the path matching benchmark includes: generating a node benchmark record for each technical node, wherein the node benchmark record includes the node technical objective, node constraints, parameter allowable range, data source requirements, method applicability conditions, and verification pass conditions; binding the node benchmark record with the edge object in the R&D path relationship chain to form a benchmark set for judging the matching relationship between technical nodes; the benchmark set includes target matching benchmark, constraint matching benchmark, parameter matching benchmark, data source matching benchmark, method matching benchmark, and verification result matching benchmark.

[0009] The present invention is further configured such that the monitoring of path change events in the target R&D path includes: receiving change logs or change messages of the target R&D path and generating path change event records; the path change event records include event type, change element identifier, element value before change, element value after change, trigger time, and trigger source; querying the R&D path element set according to the change element identifier to determine the changed technical node corresponding to the path change event; performing node tracking along the node relationships connected to the changed technical node in the R&D path relationship chain to obtain the node influence domain; the node influence domain is composed of a set of technical nodes affected by the path change event.

[0010] The present invention is further configured such that determining the conflict propagation path includes: taking the changed technology node as the propagation starting point, searching for associated technology nodes according to the tracing direction corresponding to the node relationship type; recording the sequence of edge objects traversed between the changed technology node and each associated technology node; determining the sequence of edge objects as the conflict propagation path; the conflict propagation path includes a propagation starting point, a propagation ending point, the technology nodes traversed, the node relationship type traversed, and the node association level; the node association level is the number of edge objects traversed between the changed technology node and the propagation ending point.

[0011] The present invention is further configured such that the generation of the node conflict factor includes: comparing the changed element value corresponding to the path change event, the technical nodes in the node influence domain, and the node relationships in the conflict propagation path with the path matching benchmark according to the conflict judgment rules corresponding to the node relationship type; the comparison is performed by generating corresponding node conflict factors based on output result consistency, constraint condition satisfaction, substitution condition consistency, mutual exclusion condition triggering, or verification result acceptance consistency when the node relationship type is a dependency relationship, constraint relationship, substitution relationship, mutual exclusion relationship, or verification acceptance relationship; the node conflict factor includes a conflict factor identifier, conflict type, conflict technical node identifier, changed element identifier, broken matching benchmark, conflict degree value, and node association level; the R&D path conflict value is determined based on the conflict degree value of the node conflict factor, the node relationship type weight, the node association level attenuation coefficient, and the number of conflict propagation paths. The R&D path conflict value is calculated according to the following formula: Where Di is the conflict degree value of the conflict factor of the i-th node, Wi is the conflict type weight corresponding to the conflict factor of the i-th node, Pi is the node relationship type weight corresponding to the conflict factor of the i-th node, and Qi is the node association level attenuation coefficient corresponding to the conflict factor of the i-th node; the node association level attenuation coefficient Li represents the node association level of the i-th node conflict factor; when the R&D path conflict value is not less than the preset warning threshold, it is determined that the path change event disrupts the target R&D path matching relationship.

[0012] The present invention is further configured such that the technical conflict early warning result includes the conflicting technical node, the conflict source, the damaged matching benchmark, the conflict propagation path, the R&D path conflict value, and the early warning level; the early warning level is determined based on the R&D path conflict value and threshold table; after receiving the confirmed conflict handling result, the conflict handling result is written into the corresponding node benchmark record, and the relationship constraint expression, node relationship type weight, or relationship direction of the corresponding edge object in the R&D path relationship chain is updated.

[0013] The beneficial effects of this invention are as follows: By extracting and unifying the elements of the target R&D path data, this invention constructs an R&D path relationship chain that includes dependency relationships, constraint relationships, substitution relationships, mutual exclusion relationships, and verification succession relationships. Combined with path matching benchmarks, it locates path change events, tracks the impact domain, and analyzes conflict propagation paths. This enables real-time identification and early warning of the disruption of technical node matching relationships in the R&D path. It can promptly detect technical conflicts caused by changes in objectives, constraints, parameters, data sources, method selection, or verification results during the R&D process, thereby improving the timeliness and accuracy of R&D path risk identification.

[0014] This invention uses node conflict factors and R&D path conflict values ​​to quantitatively evaluate the degree of conflict, and determines the warning level based on the conflict propagation path, node relationship type weight, and node association hierarchy attenuation coefficient. This serves to classify and judge the scope and severity of the impact of technical conflicts. At the same time, by updating the path matching benchmark and R&D path relationship chain based on the confirmed conflict handling results, the basis for warning analysis can be dynamically corrected as the R&D process progresses, improving the adaptability and reliability of subsequent conflict identification, thereby reducing the technical risks in R&D path adjustments. Attached Figure Description

[0015] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below.

[0016] Figure 1 This is a flowchart of the overall method of the present invention.

[0017] Figure 2 This is a flowchart of the node conflict factor generation process of the present invention.

[0018] Figure 3 This is a flowchart of the early warning and benchmark update process of this invention. Detailed Implementation

[0019] The technical solutions of the present invention will be described below with reference to the accompanying drawings. The described embodiments are only some embodiments of the present invention, and not all embodiments.

[0020] Example 1 Please see Figures 1-3 This invention provides a real-time early warning and analysis method for technology conflicts in R&D paths, comprising: S1. Obtain the R&D path data for the target R&D path, extract and unify the elements of the R&D path data, and generate an R&D path element set. The R&D path element set includes execution status elements and matching relationship elements. Execution status elements include technical node identifiers, node execution status, node version numbers, and node timing markers. Matching relationship elements include technical target items, technical constraint items, parameter condition items, data source items, method selection items, and verification result items. Element extraction and unification of identification includes: identifying execution status elements and matching relationship elements from the R&D path data, generating element records for each R&D path element, and merging different names, different format parameters, and different version identifiers representing the same technical object under the same standard identifier.

[0021] S2. Construct an R&D path relationship chain based on the R&D path element set. The R&D path relationship chain includes node relationships used to characterize the matching relationships between technical nodes. Map the R&D path elements in the R&D path element set to the corresponding technical nodes to generate path matching benchmarks. Node relationships include dependency relationships, constraint relationships, substitution relationships, mutual exclusion relationships, and verification succession relationships. The R&D path relationship chain is a directed relationship graph with technical nodes as node objects and node relationships as edge objects. Edge objects include the starting technical node identifier, the ending technical node identifier, the node relationship type, the relationship direction, and the relationship constraint expression. The relationship constraint expression is used to record the matching conditions between two technical nodes on technical objective items, technical constraint items, parameter condition items, data source items, method selection items, or verification result items. The generation of path matching benchmarks includes: generating a node benchmark record for each technical node. The node benchmark record includes the node technical objective, node constraint condition, parameter allowable range, data source requirements, method applicability conditions, and verification pass conditions. Bind the node benchmark record to the edge objects in the R&D path relationship chain to form a benchmark set used to determine the matching relationship between technical nodes. The benchmark set includes target matching benchmarks, constraint matching benchmarks, parameter matching benchmarks, data source matching benchmarks, method matching benchmarks, and verification result matching benchmarks.

[0022] S3. Monitor path change events in the target R&D path, locate the path change events to the corresponding changed technical nodes, and determine the node influence domain and conflict propagation path of the changed technical nodes based on the R&D path relationship chain. Monitoring path change events in the target R&D path includes: receiving change logs or change messages of the target R&D path and generating path change event records. Path change event records include event type, change element identifier, element value before change, element value after change, trigger time, and trigger source. Query the R&D path element set based on the change element identifier to determine the changed technical node corresponding to the path change event. Track nodes along the node relationships connected to the changed technical node in the R&D path relationship chain to obtain the node influence domain. The node influence domain consists of the set of technical nodes affected by the path change event. Determining the conflict propagation path includes: using the changed technical node as the propagation starting point, searching for related technical nodes according to the tracing direction corresponding to the node relationship type. Record the sequence of edge objects traversed between the changed technical node and each related technical node. Determine the conflict propagation path as the conflict propagation path. The conflict propagation path includes the propagation starting point, propagation ending point, traversed technical nodes, traversed node relationship types, and node association levels. The node association level is the number of edge objects traversed between the change technology node and the propagation endpoint.

[0023] S4. Compare the path change event, node influence domain, and conflict propagation path with the path matching benchmark according to the conflict determination rules corresponding to the node relationships to generate node conflict factors. Calculate the R&D path conflict value based on the node conflict factors. The R&D path conflict value characterizes the degree to which the path change event disrupts the target R&D path matching relationship. The generation of node conflict factors includes: comparing the changed element values ​​corresponding to the path change event, the technical nodes in the node influence domain, and the node relationships in the conflict propagation path with the path matching benchmark according to the conflict determination rules corresponding to the node relationship type. For node relationship types such as dependency, constraint, substitution, mutual exclusion, or verification continuation, corresponding node conflict factors are generated based on output consistency, constraint satisfaction, substitution consistency, mutual exclusion triggering, or verification result continuation consistency, respectively. Node conflict factors include conflict factor identifier, conflict type, conflict technical node identifier, changed element identifier, disrupted matching benchmark, conflict degree value, and node association level. The R&D path conflict value is determined based on the conflict degree value of the node conflict factor, the node relationship type weight, the node association level attenuation coefficient, and the number of conflict propagation paths. The R&D path conflict value is calculated using the following formula: Where Di is the conflict degree value of the conflict factor of the i-th node, Wi is the conflict type weight corresponding to the conflict factor of the i-th node, Pi is the node relationship type weight corresponding to the conflict factor of the i-th node, and Qi is the node association level attenuation coefficient corresponding to the conflict factor of the i-th node. Li represents the node association level of the i-th node conflict factor. When the R&D path conflict value is not less than the preset warning threshold, the path change event is determined to have disrupted the target R&D path matching relationship.

[0024] S5. When the R&D path conflict value reaches the preset warning threshold, a technical conflict warning result is generated, and the path matching benchmark and R&D path relationship chain are updated based on the confirmed conflict resolution results. The technical conflict warning result includes the conflicting technical node, conflict source, disrupted matching benchmark, conflict propagation path, R&D path conflict value, and warning level. The warning level is determined based on the R&D path conflict value and threshold table. After receiving the confirmed conflict resolution result, the conflict resolution result is written to the corresponding node benchmark record, and the relationship constraint expression, node relationship type weight, or relationship direction of the corresponding edge object in the R&D path relationship chain is updated.

[0025] This invention enables the analysis of changes in the R&D path by establishing corresponding relationships between technical nodes. This allows for the determination of the impact of changes in technical objectives, constraints, parameters, data sources, method selection, and verification results on the matching status of relevant technical nodes, and further identifies the location, scope, and propagation direction of conflicts. Consequently, technical conflicts in the R&D path are no longer merely identified as single-point anomalies, but rather as correlation analysis results based on node relationships and matching conditions.

[0026] Meanwhile, this invention can transform conflict identification results into early warning information that includes conflict source, affected nodes, propagation path and conflict degree, and update the matching benchmark and path relationship based on the confirmed processing results, realizing the connection between R&D path change, conflict judgment, early warning output and rule correction, so that R&D path technical conflict analysis has dynamic response and continuous correction capabilities.

[0027] Example 2 Please see Figure 2 Based on Example 1, this example takes the development path of a liquid-cooled heat dissipation base plate for a certain type of high-power electronic control unit as the object, and illustrates the process of handling the early warning of technical conflicts in the development path caused by changes in development cost constraints.

[0028] 1. Research and Development Path and Initial Matching Benchmark The research and development path includes aluminum alloy substrate selection node N1, microchannel structure design node N2, prototype processing and welding node N3, thermal cycling reliability verification node N4, and complete machine assembly and matching node N5. Among them, N3, N4, and N5 correspond to prototype processing, reliability verification, and assembly matching, respectively, and are the conflict analysis nodes in this embodiment.

[0029] In the R&D path relationship chain, N2 to N3 represent a dependency relationship, N3 to N4 represent a verification succession relationship, and N4 to N5 represent a constraint relationship. The system generates path matching benchmarks based on the R&D path data, which mainly include: the sample processing cost for node N3 must not be less than 300,000 yuan, and at least 10 samples should be produced; the verification cost for node N4 must not be less than 380,000 yuan, and at least 300 thermal cycle verifications should be completed; and node N5 should form an assembly matching result based on at least 8 valid samples.

[0030] The aforementioned R&D expenses are not separate financial budget data, but rather technical constraints tied to the number of prototypes, processing accuracy, number of verifications, and assembly verification conditions.

[0031] 2. Determination of Route Change Events and Impact Areas During the R&D execution phase, the system detected a path change event E20260318-01, with the event type being R&D expense adjustment.

[0032] The cost adjustments are as follows: the sample processing cost for node N3 is reduced from 320,000 yuan to 240,000 yuan, and the thermal cycling verification cost for node N4 is reduced from 410,000 yuan to 230,000 yuan. The system locates the event to the two changed technical nodes, N3 and N4, based on the change element identifiers N3-FEE and N4-FEE.

[0033] Subsequently, the system traces nodes along the R&D path relationship chain, obtaining node influence domains of N3, N4, and N5. The corresponding conflict propagation paths include N3-N4, N4-N5, and N3-N4-N5. Specifically, N3-N4 corresponds to a verification acceptance relationship with a node association level of 1; N4-N5 corresponds to a constraint relationship with a node association level of 1; and N3-N4-N5 is a composite propagation path formed by consecutive verification acceptance and constraint relationships with a node association level of 2.

[0034] 3. Generation of node conflict factors N3 Node Cost Constraint Conflict Factor: The path matching benchmark for N3 node is that the sample processing cost shall not be less than 300,000 yuan, and the number of samples shall not be less than 10. After cost adjustment, the available cost for N3 node is 240,000 yuan.

[0035] According to the sample processing quotation, 240,000 yuan can only complete the processing of 6 samples and the retesting of 4 of them, which is lower than the benchmark requirement of no less than 10 samples for node N3. Among them, the effective samples that can be used for N5 assembly matching are counted as 6 samples that have been processed and passed the initial inspection.

[0036] Therefore, node N3 does not meet the sample processing cost benchmark and sample quantity benchmark. The system generates a node conflict factor C1, the conflict type is cost constraint conflict, the conflicting technical node is N3, the change element identifier is N3-FEE, and the broken matching benchmark is that the sample processing cost shall not be less than 300,000 yuan.

[0037] The conflict level value D1 of C1 is determined according to the proportion of the cost gap: N4 node verification conflict factor: The path matching benchmark for N4 nodes is that the verification cost shall not be less than 380,000 yuan, and at least 300 hot cycles shall be completed. After the cost adjustment, the available cost for N4 nodes is 230,000 yuan.

[0038] According to the test quotation, the basic cost of the thermal cycling test is 150,000 yuan, and the thermal cycling execution cost is 500 yuan per cycle. After deducting the basic cost, the total cost of 230,000 yuan can support 160 thermal cycles, which is lower than the baseline requirement of 300 thermal cycles for the N4 node.

[0039] Therefore, node N4 cannot generate a thermal cycle verification result that satisfies the path matching benchmark and cannot be used as a valid input for assembly matching at node N5. The system generates a node conflict factor C2, with the conflict type being verification acceptance conflict, the conflicting technical node being N4, the change element identifier being N4-FEE, and the violated matching benchmark being at least 300 thermal cycles.

[0040] The conflict level value D2 of C2 is determined according to the gap ratio of the number of verifications: N5 node assembly constraint conflict factor: The N5 node requires an assembly matching result based on at least 8 valid samples. Since the N3 node, after cost adjustments, is expected to generate only 6 samples, this is lower than the N5 node's assembly verification input requirement.

[0041] The system generates a node conflict factor of C3, the conflict type is constraint relationship conflict, the conflict technology node is N5, the change element identifier is N3-FEE, and the broken matching benchmark is that the number of valid samples is not less than 8.

[0042] The conflict level value D3 of C3 is determined according to the proportion of the sample quantity shortage: 4. Calculation of R&D path conflict value The system calculates the development path conflict value R according to the following formula: Where Di is the conflict degree value of the i-th node conflict factor, Wi is the conflict type weight corresponding to the i-th node conflict factor, Pi is the node relationship type weight corresponding to the i-th node conflict factor, and Qi is the node association hierarchy attenuation coefficient corresponding to the i-th node conflict factor.

[0043] Node association hierarchy attenuation coefficient: .

[0044] in, The preset attenuation constant is set to 0.35 in this embodiment, and Li is the node association level of the i-th node conflict factor.

[0045] The weights for conflict types and node relationship types are determined by a pre-set weight table. The parameter table is configured based on records of prototype scrapping, verification termination, and assembly rework in historical R&D projects, and can be confirmed by the project technical lead during the system initialization phase.

[0046] Among them, the weight of cost constraint conflict is 0.80, the weight of verification acceptance conflict is 0.95, the weight of constraint relationship conflict is 0.85, the weight of verification acceptance relationship is 0.90, the weight of constraint relationship is 0.85, and the weight of node relationship type corresponding to cost constraint conflict is taken according to the weight of constraint relationship, which is 0.85.

[0047] Table 1: Parameters for calculating node conflict factor.

[0048] According to Table 1, the R&D path conflict value R is: 5. Early warning result generation The warning thresholds are set as follows: no warning is issued when R < 0.30, a general warning is generated when 0.30 ≤ R < 0.45, a level 2 warning is generated when 0.45 ≤ R < 0.65, and a level 1 warning is generated when R ≥ 0.65.

[0049] The calculated R&D path conflict value R is 0.503, falling within the Level 2 warning range. The conflict source corresponding to the warning result is path change event E20260318-01. The conflicting technical nodes include N3, N4, and N5. The disrupted matching benchmarks include the sample processing cost benchmark, the thermal cycle verification number benchmark, and the assembly sample quantity benchmark. The conflict propagation paths include N3-N4, N4-N5, and N3-N4-N5. The warning level is Level 2.

[0050] 6. Path matching baseline update After confirming the warning, the project team provided the following solutions: retain the requirement of 300 thermal cycles for verification at node N4, reduce the number of simulation rounds for optimizing non-critical structures at node N2 from 6 to 4, and adjust the sample processing method for node N3 to two batches.

[0051] After a revised quotation, the cost of node N3 has been adjusted to 290,000 yuan, and the cost of node N4 has been adjusted to 350,000 yuan. The system's pre-set quotation parameter table shows that, under the condition of using shared fixture scheduling, the 350,000 yuan quotation corresponds to a thermal cycling verification capability of 300 cycles.

[0052] Table 2: Path matching baseline update table.

[0053] In summary, the system transforms changes in R&D costs into changes in technical constraints in the R&D path, and accordingly completes the identification of cost change events, determination of node impact domains, generation of conflict factors, calculation of conflict values, output of early warnings, and updating of path matching benchmarks, which is used to determine technical conflicts between prototype processing, reliability verification, and assembly matching.

[0054] Example 3 Please see Figure 3 Based on Examples 1 and 2, this example takes the research and development path of the thermal management control strategy for low-temperature fast charging of battery packs as the object, and explains how the system identifies and quantifies the degree of damage to the subsequent technical node matching relationship based on the node influence domain, conflict propagation path and path matching benchmark when the method selection options change.

[0055] 1. Research and Development Path and Baseline Configuration The R&D path includes five technical nodes: data acquisition node T01, thermal model construction node T02, control strategy generation node T03, bench verification node T04, and overall package safety assessment node T05.

[0056] In the initial state, node T02 adopts the M1 method, which is a physical constraint parameter identification method. Its input conditions are: no less than 60 test cycles, ambient temperature covering 0℃ to 40℃, sampling interval no greater than 1s, and data integrity rate no less than 90%. The output results of M1 include thermal resistance parameter R, heat capacity parameter C, and temperature rise prediction curve.

[0057] The control strategy for node T03 uses the thermal resistance parameter R, thermal capacity parameter C, and temperature rise prediction slope output from T02 as inputs to determine the fast charging derating point. The bench verification scheme for node T04 uses the temperature rise prediction curve output by method M1 as the verification object. The verification pass conditions are: the maximum temperature prediction error is no greater than 1.8℃, the derating trigger time is no greater than 8s, and the maximum temperature difference within the battery pack is no greater than 5℃.

[0058] In the R&D path relationship chain, T01 and T02 have a dependency relationship, T02 and T03 have a substitution relationship, and node T03 has a method applicability constraint on the method output result of node T02. T03 and T04 have a verification succession relationship, and T04 and T05 have a verification succession relationship.

[0059] Among them, the data source matching benchmark corresponds to the input data conditions of T01 to T02, the method matching benchmark corresponds to the output fields and method applicable conditions of T02 to T03, and the verification result matching benchmark corresponds to the verification objects, verification indicators and evaluation input conditions of T03 to T04 and T04 to T05.

[0060] 2. Path change event The system detected a change in the method selection option at node T02. The event type of the path change event is method selection option change, the change element identifier is T02-METHOD, the element value before the change is M1: physical constraint parameter identification method, the element value after the change is M2: data-driven gradient boosting prediction method, the trigger time is 14:26 on July 18, 2025, and the trigger source is the thermal management algorithm group version submission record.

[0061] The M2 method is applicable under the following conditions: the training data must be looped at least 240 times, the ambient temperature must cover -20℃ to 50℃, the sampling interval must be no greater than 0.5s, and the data integrity rate must be no less than 95%. Each set of training data must also include five synchronous fields: cell surface temperature, current, voltage, ambient temperature, and coolant inlet temperature. The output of the M2 method is the predicted temperature value after 30s, 60s, and 120s, and it does not directly output the thermal resistance parameter R and the heat capacity parameter C.

[0062] The system locates the change event to node T02 and determines the node's impact domain as T01, T03, T04, and T05. The corresponding conflict propagation paths are: T02-T01, node relationship type is dependency, node association level is 1; T02-T03, node relationship type is substitution, node association level is 1; T02-T03-T04, node relationship types include substitution and verification acceptance, node association level is 2; T02-T03-T04-T05, node relationship types include substitution and verification acceptance, node association level is 3.

[0063] 3. Generation of node conflict factors The system compares the applicable conditions of the M2 method with the path matching benchmark in the node influence domain.

[0064] Data source: Conflict factor F1: The current test data for T01 is as follows: 86 test cycles, ambient temperature range from 0℃ to 35℃, sampling interval of 1s, and data integrity rate of 93.2%. Among the data records corresponding to the 86 test cycles, the coolant inlet temperature field is missing in 18 sets.

[0065] The above data does not meet the requirements of the M2 method for training data cycle count, ambient temperature range, sampling interval, data integrity rate, and field integrity. Based on this, the system generates a data source conflict factor F1. The conflict technical node of F1 is T01, and the broken matching benchmark is the data source matching benchmark. Since the current data of T01 has five types of deviations, namely insufficient cycle count, insufficient ambient temperature coverage, unsatisfactory sampling interval, insufficient data integrity rate, and missing coolant inlet temperature field, the conflict degree value D1 is set to 0.62.

[0066] Method to replace conflict factor F2: The original control strategy of T03 requires reading the thermal resistance parameter R, the heat capacity parameter C, and the temperature rise prediction slope. The M2 method only outputs the temperature prediction values ​​at multiple time points, and does not output the thermal resistance parameter R and the heat capacity parameter C.

[0067] Therefore, the original control strategy of T03 cannot directly accept the new output result of T02. The system generation method replaces the conflict factor F2. The conflict technical node of F2 is T03, and the broken matching benchmark is the method matching benchmark. Since the M2 method lacks the thermal resistance parameter R and thermal capacity parameter C required by the original control strategy of T03, the conflict degree value D2 is taken as 0.75.

[0068] Verify the conflict factor F3: The current bench validation scheme for T04 is numbered V-TM-2025-04, and the validation object is the temperature rise prediction curve output by the M1 method. This scheme does not set up test items for prediction error under operating conditions outside the training set, prediction error in the low temperature range, and sensor missing stability, nor does it retain the cross-validation dataset required by the M2 method.

[0069] Therefore, the verification scheme of T04 cannot directly accept the M2 method. The system generates a verification acceptance conflict factor F3. The conflicting technical node of F3 is T04. The broken matching benchmark is the verification result matching benchmark. Since the current verification scheme lacks three types of M2 method verification items and does not retain the cross-validation dataset, the conflict degree value D3 is 0.66.

[0070] The verification results follow the conflict factor F4: The upper limit of error referenced in the T05 overall safety assessment node is 1.8℃, which is derived from the bench verification results corresponding to the M1 method. Since T04 has not yet completed the verification for the M2 method, the current safety assessment input for T05 does not have a valid basis for acceptance.

[0071] The system generates a verification result that carries over conflict factor F4. The conflict technical node of F4 is T05, and the broken matching benchmark is the verification result matching benchmark. The conflict degree value D4 is 0.58.

[0072] For F4, although node T05 does not directly receive the method output from T02, its security assessment is based on the verification results from T04. Since T04 has not yet completed verification for the M2 method, the evaluation input for T05 carries an indirect risk of failure. Based on the downstream verification result scoring rules, D4 is assigned a score of 0.58.

[0073] 4. Calculation of R&D path conflict value The node association hierarchy attenuation coefficient Qi is determined by the following formula: Where Li represents the node association level.

[0074] Table 1: Parameters for calculating node conflict factor.

[0075] The R&D path conflict value R is calculated as follows: 。

[0076] The calculated value is R = 1.57.

[0077] The warning level is determined according to the range of the R&D path conflict value R: when R < 0.80, it is a low-level warning; when 0.80 ≤ R < 1.20, it is a medium-level warning; and when R ≥ 1.20, it is a high-level warning. The preset warning threshold is 1.20. Since the R&D path conflict value of 1.57 is greater than the preset warning threshold of 1.20, the system determines that the method selection change of node T02 has disrupted the matching relationship of this R&D path.

[0078] 5. Early warning results and route updates The system generates a technical conflict warning result. The conflict source is the method selection option of node T02, which has changed from M1 to M2. The conflicting technical nodes are T01, T03, T04, and T05. The broken matching benchmarks are the data source matching benchmark, the method matching benchmark, and the verification result matching benchmark. The conflict propagation paths are T02-T01, T02-T03, T02-T03-T04, and T02-T03-T04-T05. The R&D path conflict value is 1.57, and the warning level is a high-level warning.

[0079] After the conflict resolution results are confirmed, the system updates the corresponding node baseline records and R&D path relationship chains. The data acquisition baseline for T01 is updated to: no less than 240 test cycles, sampling interval no greater than 0.5s, data integrity rate no less than 95%, and a coolant inlet temperature field is added. The method applicability conditions for T03 are updated to: able to receive 30s, 60s, and 120s temperature prediction values ​​output by M2. The verification scheme for T04 adds low-temperature range prediction error, training set out-of-set condition prediction error, and sensor missing stability test items. The edge object relationship constraint expressions from T02 to T03 are updated from output thermal resistance parameter R and thermal capacity parameter C to output multi-time temperature prediction values ​​and provide error confidence intervals.

[0080] Therefore, after the method selection options change, the system can determine the affected technical nodes based on the node influence domain and conflict propagation path, and determine the degree of damage to the R&D path matching relationship caused by the change based on the node conflict factor and R&D path conflict value.

[0081] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A real-time early warning analysis method for R&D path technology conflicts, characterized in that, include: S1. Obtain the R&D path data of the target R&D path, extract elements and unify the identification of the R&D path data, and generate a R&D path element set; The R&D path element set includes execution status elements and matching relationship elements; S2. Construct a research and development path relationship chain based on the research and development path element set, wherein the research and development path relationship chain includes node relationships used to characterize the matching relationship between technology nodes; Map the R&D path elements in the R&D path element set to the corresponding technology nodes to generate a path matching benchmark; S3. Monitor path change events in the target R&D path, locate the path change events to the corresponding changed technology nodes, and determine the node influence domain and conflict propagation path of the changed technology nodes based on the R&D path relationship chain; S4. Compare the path change event, the node influence domain, and the conflict propagation path with the path matching benchmark according to the conflict determination rules corresponding to the node relationship to generate a node conflict factor; The R&D path conflict value is calculated based on the node conflict factor, and the R&D path conflict value characterizes the degree to which the path change event disrupts the target R&D path matching relationship; S5. When the R&D path conflict value reaches the preset warning threshold, generate a technical conflict warning result, and update the path matching benchmark and the R&D path relationship chain according to the confirmed conflict handling result.

2. The real-time early warning analysis method for R&D path technology conflicts according to claim 1, characterized in that: The execution status elements include technical node identifier, node execution status, node version number, and node timing marker; The matching relationship elements include technical objective items, technical constraint items, parameter condition items, data source items, method selection items, and verification result items; The unified extraction and identification of elements includes: identifying the execution status elements and the matching relationship elements from the R&D path data, generating element records for each R&D path element, and merging different names, different format parameters, and different version identifiers representing the same technical object under the same standard identifier.

3. The real-time early warning analysis method for R&D path technology conflicts according to claim 1, characterized in that: The node relationships include dependency relationships, constraint relationships, substitution relationships, mutual exclusion relationships, and verification succession relationships; the R&D path relationship chain is a directed relationship graph with technical nodes as node objects and the node relationships as edge objects; the edge objects include the starting technical node identifier, the ending technical node identifier, the node relationship type, the relationship direction, and the relationship constraint expression; the relationship constraint expression is used to record the matching conditions between two technical nodes on technical target items, technical constraint items, parameter condition items, data source items, method selection items, or verification result items.

4. The real-time early warning analysis method for R&D path technology conflicts according to claim 1, characterized in that: The generation of the path matching benchmark includes: generating a node benchmark record for each technical node, wherein the node benchmark record includes the node's technical objectives, node constraints, parameter allowable ranges, data source requirements, method applicability conditions, and verification pass conditions; binding the node benchmark record with the edge objects in the R&D path relationship chain to form a benchmark set for judging the matching relationship between technical nodes; the benchmark set includes target matching benchmarks, constraint matching benchmarks, parameter matching benchmarks, data source matching benchmarks, method matching benchmarks, and verification result matching benchmarks.

5. The real-time early warning analysis method for R&D path technology conflicts according to claim 1, characterized in that: The monitoring of path change events in the target R&D path includes: receiving change logs or change messages of the target R&D path and generating path change event records; the path change event records include event type, change element identifier, element value before change, element value after change, trigger time, and trigger source; querying the R&D path element set according to the change element identifier to determine the changed technical node corresponding to the path change event; tracing nodes along the node relationships connected to the changed technical node in the R&D path relationship chain to obtain the node influence domain; the node influence domain is composed of a set of technical nodes affected by the path change event.

6. The real-time early warning analysis method for R&D path technology conflicts according to claim 3, characterized in that: The determination of the conflict propagation path includes: taking the changed technology node as the propagation starting point, searching for associated technology nodes according to the tracing direction corresponding to the node relationship type; recording the sequence of edge objects traversed between the changed technology node and each associated technology node; and determining the sequence of edge objects as the conflict propagation path. The conflict propagation path includes a propagation starting point, a propagation ending point, the technology nodes traversed, the node relationship type traversed, and the node association level. The node association level is the number of edge objects traversed between the changed technology node and the propagation ending point.

7. The real-time early warning analysis method for R&D path technology conflicts according to claim 6, characterized in that: The generation of the node conflict factor includes: comparing the changed element value corresponding to the path change event, the technical nodes in the node influence domain, and the node relationships in the conflict propagation path with the path matching benchmark according to the conflict judgment rules corresponding to the node relationship type; the comparison is to generate corresponding node conflict factors based on the consistency of output results, the satisfaction of constraint conditions, the consistency of substitution conditions, the triggering of mutual exclusion conditions, or the consistency of verification results when the node relationship type is a dependency relationship, constraint relationship, substitution relationship, mutual exclusion relationship, or verification acceptance relationship; the node conflict factor includes a conflict factor identifier, conflict type, conflict technical node identifier, changed element identifier, broken matching benchmark, conflict degree value, and node association level; the R&D path conflict value is determined based on the conflict degree value of the node conflict factor, the weight of the node relationship type, the attenuation coefficient of the node association level, and the number of conflict propagation paths.

8. The real-time early warning analysis method for R&D path technology conflicts according to claim 7, characterized in that: The technical conflict early warning results include conflicting technical nodes, conflict sources, disrupted matching benchmarks, conflict propagation paths, R&D path conflict values, and early warning levels. The warning level is determined based on the R&D path conflict value and threshold table; After receiving the confirmed conflict resolution result, the conflict resolution result is written into the corresponding node baseline record, and the relationship constraint expression, node relationship type weight or relationship direction of the corresponding edge object in the R&D path relationship chain are updated.

Images