Diesel engine digital health diagnosis method based on multi-source data fusion

Abstract

This invention discloses a digital health diagnosis method for diesel engines based on multi-source data fusion, comprising the following steps: collecting signals generated during diesel engine operation and preprocessing them to generate multi-source combustion response data sequences; mapping events to combustion propagation nodes to generate a set of combustion propagation nodes; establishing propagation connections between nodes to generate a set of combustion propagation paths; constructing a combustion propagation topology; calculating the frequency of combustion propagation path occurrences and the stability of propagation time differences to generate a stable domain for the combustion propagation structure; performing structural matching between the current combustion propagation topology and the stable domain for the combustion propagation structure to generate a topology deviation structure; determining the diesel engine health status based on the topology deviation structure and generating a digital health diagnosis result for the diesel engine. This invention utilizes multi-source data fusion and combustion propagation topology analysis methods to achieve abnormal identification of diesel engine combustion structures, possessing advantages such as high diagnostic accuracy and strong structural characterization capabilities.

Description

Technical Field

[0001] This invention relates to the field of digital health diagnosis of diesel engines, and more particularly to a method for digital health diagnosis of diesel engines based on multi-source data fusion. Background Technology

[0002] Diesel engines are widely used in engineering machinery, marine power and power generation equipment. Their operating status directly affects the reliability and service life of the equipment. In the existing technology, the monitoring of the health status of diesel engines usually relies on single or a small number of sensor signals, such as cylinder pressure signals, vibration signals or temperature signals. The operating status is evaluated by means of threshold judgment, feature statistics or spectrum analysis to identify abnormal combustion or fault conditions. These methods can reflect some operating characteristics in practical applications, but they do not make sufficient use of the correlation between multi-source combustion response signals.

[0003] The combustion process of diesel engines exhibits significant temporal propagation characteristics, with complex propagation relationships between different combustion response events. Relying solely on a single signal feature or simple statistical indicators is insufficient to accurately characterize changes in the combustion propagation structure. When anomalies occur in the combustion process, existing technologies struggle to analyze the combustion propagation relationship at the overall structural level, resulting in low accuracy in anomaly identification and difficulty in effectively characterizing the location and structural changes of combustion propagation anomalies. Therefore, it is necessary to propose a digital health diagnosis method for diesel engines that can construct a combustion propagation topology based on multi-source data and perform structural deviation analysis. Summary of the Invention

[0004] One objective of this invention is to propose a digital health diagnosis method for diesel engines based on multi-source data fusion. This invention utilizes multi-source data fusion and combustion propagation topology analysis to identify abnormalities in the combustion structure of diesel engines, and has the advantages of high diagnostic accuracy and strong structural characterization capability.

[0005] The digital health diagnosis method for diesel engines based on multi-source data fusion according to embodiments of the present invention includes the following steps: Collect cylinder pressure signals, engine vibration signals, crankshaft speed signals, exhaust temperature signals, and fuel injection trigger signals generated during diesel engine operation, and preprocess them to generate multi-source combustion response data sequences; Based on the diesel engine combustion cycle, the multi-source combustion response data sequence is divided into time windows. Within each combustion cycle, cylinder pressure peak events, vibration response peak events, exhaust temperature rise events, and fuel injection trigger events are identified. These events are mapped to combustion propagation nodes, generating a set of combustion propagation nodes. Based on the time index relationship of each node in the combustion propagation node set, establish the propagation connection relationship between the nodes and generate a set of combustion propagation paths; Construct a combustion propagation topology based on the set of combustion propagation paths; Statistical analysis of the combustion propagation topology is performed over multiple consecutive combustion cycles to calculate the frequency of combustion propagation path occurrence and the stability of propagation time difference, thereby generating a stable domain for the combustion propagation structure. The current combustion propagation topology is matched with the stability domain of the combustion propagation structure, the topology deviation determination quantity is calculated, and the topology deviation structure is generated. The health status of the diesel engine is determined based on the topological deviation structure, and a digital health diagnosis result for the diesel engine is generated.

[0006] Optionally, the preprocessing specifically includes data synchronization calibration, noise filtering, outlier removal, data resampling, signal normalization, and time series alignment.

[0007] Optionally, the generation of the combustion propagation node set specifically includes: The multi-source combustion response data sequence is divided into cycles according to the crankshaft speed signal. The start and end positions of each combustion cycle are determined based on the crankshaft speed change. The multi-source combustion response data sequence is then segmented into time windows to generate multiple combustion cycle time window sequences. The cylinder pressure signal is scanned within each combustion cycle time window to identify the location of the local maximum value of the cylinder pressure signal, and the time point corresponding to the location is determined as the cylinder pressure peak event, generating a cylinder pressure peak event sequence. Within the same combustion cycle time window, the vibration signal of the engine body is scanned, the peak position of the vibration amplitude of the vibration signal is identified, the time point corresponding to the position is determined as the vibration response peak event, and a vibration response peak event sequence is generated. Within the combustion working cycle time window, the exhaust temperature signal is continuously detected to identify the rapid rise range of exhaust temperature. The position where the exhaust temperature change rate reaches the preset threshold is determined as the exhaust temperature rise event, and an exhaust temperature rise event sequence is generated. The fuel injection trigger signal is read within the combustion working cycle time window, and the fuel injection trigger event is identified according to the trigger time of the fuel injection trigger signal, and a fuel injection trigger event sequence is generated. A unified time index is calibrated for the cylinder pressure peak event sequence, vibration response peak event sequence, exhaust temperature rise event sequence, and fuel injection trigger event sequence. Various events are recorded according to their occurrence time to generate a combustion event time index set. The various events in the combustion event time index set are mapped to combustion propagation nodes, and the combustion propagation nodes are categorized and organized according to the combustion work cycle to generate a combustion propagation node set.

[0008] Optionally, the generation of the set of combustion propagation paths specifically includes: Obtain the set of combustion propagation nodes, extract the time index corresponding to each combustion propagation node, and sort the combustion propagation nodes in ascending order according to the time index to generate a time-ordered sequence of combustion propagation nodes; Based on the time-ordered combustion propagation node sequence, the time difference between adjacent combustion propagation nodes in the sequence is calculated, and the node time interval information between each combustion propagation node is recorded to generate a node time interval sequence. The propagation relationship between each combustion propagation node is determined based on the node time interval sequence. When the node time interval is less than the combustion propagation time threshold, a propagation connection relationship is established between the corresponding combustion propagation nodes, and a set of propagation connection edges is generated. The connection relationships between combustion propagation nodes are organized according to the propagation connection edge set. Combustion propagation nodes connected sequentially according to the propagation connection relationship are combined to form a node propagation sequence, generating a node propagation sequence set. The propagation order of the node propagation sequence set is detected, and the node propagation sequence that satisfies the increasing relationship of node time index is determined as the combustion propagation path. Each combustion propagation path is numbered to generate a set of combustion propagation paths. The set of combustion propagation paths is categorized and organized according to the combustion cycle, and the combustion propagation path corresponding to each combustion cycle is stored to generate a set of combustion propagation paths.

[0009] Optionally, the generation of the combustion propagation topology specifically includes: Each combustion propagation path in the set of combustion propagation paths is parsed according to the connection order of the combustion propagation nodes, and the sequence of combustion propagation nodes contained in each combustion propagation path is extracted to generate a set of path node sequences. Extract all combustion propagation nodes from the path node sequence set, and establish a combustion propagation node set according to the node identifier. The combustion propagation node set is then determined as the topological node set in the combustion propagation topology. Based on the node connection order in the path node sequence set, extract the node connection relationship between combustion propagation nodes, record the connection relationship between adjacent combustion propagation nodes as propagation connection edges, and generate a propagation connection edge set. Based on the set of combustion propagation nodes and the set of propagation connection edges, a node connection relationship structure is established, and the combustion propagation nodes are used as topological nodes and the propagation connection edges are used as topological edges to generate the initial combustion propagation topology structure. The frequency of each propagation connection edge in the set of combustion propagation paths is counted, and the connection frequency information corresponding to each propagation connection edge is recorded according to the frequency of occurrence, thus generating a set of propagation connection edge frequencies. Write the frequency set of propagation connection edges into the corresponding propagation connection edges in the initial combustion propagation topology to generate a combustion propagation topology containing topology nodes, propagation connection edges, and propagation connection edge frequency information.

[0010] Optionally, the generation of the combustion propagation structure stability domain specifically includes: Read the combustion propagation topology corresponding to multiple consecutive combustion working cycles, and establish a combustion propagation topology sequence according to the order of the combustion working cycles; Based on the combustion propagation topology sequence, the propagation connection edges in each combustion propagation topology are extracted, and a unified identifier is established for the propagation connection edges in different combustion work cycles to generate a set of propagation connection edges; Based on the propagation connection edge set, record the number of times each propagation connection edge appears in the combustion propagation topology sequence, and generate the propagation connection edge occurrence count set; The frequency of each propagation connection edge in the combustion working cycle sequence is calculated from the set of occurrence counts of propagation connection edges, and a set of propagation connection edge frequencies is generated. Extract the propagation time difference corresponding to each propagation connection edge in the combustion propagation topology sequence, and record the propagation time difference data of each propagation connection edge in different combustion working cycles according to the combustion working cycle order to generate a set of propagation time difference sequences; Based on the set of propagation time difference sequences, the degree of dispersion of the propagation time difference of each propagation connection edge in the combustion working cycle sequence is calculated, and a set of propagation time difference stability is generated. Based on the set of propagation connection edge frequencies and the set of propagation time difference stability, the propagation connection edges that satisfy the occurrence frequency condition and the propagation time difference stability condition are determined, and the propagation connection edges that satisfy the conditions and their corresponding combustion propagation node connection relationships are written into the combustion propagation structure stability domain to generate the combustion propagation structure stability domain.

[0011] Optionally, the generation of the topology deviation structure specifically includes: Obtain the combustion propagation topology and the stability domain of the combustion propagation structure corresponding to the current combustion cycle, and extract the propagation connection edges from the current combustion propagation topology to generate the current propagation connection edge set; Extract stable propagation connection edges from the stable domain of the combustion propagation structure to generate a set of stable propagation connection edges; Establish a correspondence between the current set of propagation connection edges and the stable set of propagation connection edges, and generate a set of propagation connection edge matching results. Based on the set of propagation connection edge matching results, identify propagation connection edges that exist in the current set of propagation connection edges but do not exist in the set of stable propagation connection edges, and generate a new set of propagation connection edges; Identify propagation connection edges that exist in the stable propagation connection edge set but do not exist in the current propagation connection edge set from the propagation connection edge matching result set, and generate a missing propagation connection edge set; The degree of structural deviation of the current combustion propagation topology from the stable domain of the combustion propagation structure is determined by the set of newly added propagation connection edges and the set of missing propagation connection edges, and a topology deviation determination quantity is generated. Based on the newly added propagation connection edge set, the missing propagation connection edge set, and the topology deviation judgment quantity, the corresponding combustion propagation node connection relationship is extracted to generate the topology deviation structure.

[0012] Optionally, the generation of the diesel engine digital health diagnosis results specifically includes: Extract the deviation propagation connection edges and the corresponding combustion propagation node connection relationships from the topological deviation structure to generate a set of deviation propagation connection edges; The number of deviation propagation connections generated in the current combustion cycle is recorded based on the set of deviation propagation connections, and the number of deviation propagation connections in each combustion cycle is recorded in the order of the combustion cycle to generate a sequence of deviation propagation connection counts. The total number of propagation connection edges in the current combustion propagation topology is extracted based on the sequence of deviation propagation connection edge counts, and the topology deviation intensity result is established based on the deviation propagation connection edge counts and the total number of propagation connection edges. Extract the combustion propagation nodes that cause structural deviations from the topological deviation structure, and record the number of times each combustion propagation node appears in the deviation propagation connection edge to generate a set of the occurrence counts of deviation propagation nodes; The combustion propagation node anomaly result is established from the set of occurrences of deviation propagation nodes, and a node propagation anomaly set is generated; An indicator for determining combustion propagation anomalies was established based on the topological deviation intensity results and the set of node propagation anomalies. The current operating status of the diesel engine is generated using the results of the combustion propagation anomaly judgment index, and the digital health diagnosis results of the diesel engine are output.

[0013] The beneficial effects of this invention are: This invention proposes a digital health diagnosis method for diesel engines based on multi-source data fusion. By collecting various operating parameters such as cylinder pressure signals, engine vibration signals, crankshaft speed signals, exhaust temperature signals, and fuel injection trigger signals, it performs unified processing on multi-source combustion response data during the diesel engine combustion process. It identifies cylinder pressure peak events, vibration response peak events, exhaust temperature rise events, and fuel injection trigger events on the combustion cycle scale, mapping different types of combustion response events as combustion propagation nodes. Furthermore, it constructs a set of combustion propagation paths and a combustion propagation topology based on the time index relationship between nodes. In this way, this invention can provide a structured expression of the diesel engine combustion process from the perspective of the propagation relationship between multi-source data. Compared with traditional diagnostic methods that rely solely on single signal features, it can more comprehensively depict the dynamic propagation laws in the combustion process, thereby improving the ability to identify complex combustion state changes.

[0014] Based on this, the present invention constructs a stable domain for combustion propagation structure by statistically analyzing the frequency of occurrence of combustion propagation paths and the stability of propagation time differences in multiple consecutive combustion working cycles. It then performs structural matching between the current combustion propagation topology and this stable domain, identifying newly added and missing propagation connection edges. Furthermore, it generates a topology deviation judgment quantity and a topology deviation structure. Through this analysis method based on topology deviation, the present invention can accurately reflect the structural change characteristics of combustion propagation relationships and effectively identify abnormal combustion propagation paths. Compared with existing technologies that rely on simple threshold judgments or feature statistics, the present invention not only improves the accuracy of diesel engine operating status diagnosis but also characterizes abnormal locations and propagation relationships from the perspective of combustion propagation structure, thereby providing more reliable technical support for the digital monitoring and fault early warning of diesel engine operating status. Attached Figure Description

[0015] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a flowchart of the diesel engine digital health diagnosis method based on multi-source data fusion proposed in this invention; Figure 2 This refers to the combustion propagation structure stability domain of the diesel engine digital health diagnosis method based on multi-source data fusion proposed in this invention. Figure 3 This is the topological deviation structure of the diesel engine digital health diagnosis method based on multi-source data fusion proposed in this invention. Detailed Implementation

[0016] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, illustrating only the basic structure of the invention, and therefore only show the components relevant to the invention.

[0017] refer to Figures 1-3 A digital health diagnosis method for diesel engines based on multi-source data fusion includes the following steps: Collect cylinder pressure signals, engine vibration signals, crankshaft speed signals, exhaust temperature signals, and fuel injection trigger signals generated during diesel engine operation, and preprocess them to generate multi-source combustion response data sequences; Based on the diesel engine combustion cycle, the multi-source combustion response data sequence is divided into time windows. Within each combustion cycle, cylinder pressure peak events, vibration response peak events, exhaust temperature rise events, and fuel injection trigger events are identified. These events are mapped to combustion propagation nodes, generating a set of combustion propagation nodes. Based on the time index relationship of each node in the combustion propagation node set, establish the propagation connection relationship between the nodes and generate a set of combustion propagation paths; Construct a combustion propagation topology based on the set of combustion propagation paths; Statistical analysis of the combustion propagation topology is performed over multiple consecutive combustion cycles to calculate the frequency of combustion propagation path occurrence and the stability of propagation time difference, thereby generating a stable domain for the combustion propagation structure. The current combustion propagation topology is matched with the stability domain of the combustion propagation structure, the topology deviation determination quantity is calculated, and the topology deviation structure is generated. The health status of the diesel engine is determined based on the topological deviation structure, and a digital health diagnosis result for the diesel engine is generated.

[0018] In this embodiment, the preprocessing specifically includes data synchronization calibration, noise filtering, outlier removal, data resampling, signal normalization, and time series alignment.

[0019] In this embodiment, the generation of the combustion propagation node set specifically includes: The multi-source combustion response data sequence is divided into cycles according to the crankshaft speed signal. The start and end positions of each combustion cycle are determined based on the crankshaft speed change. The multi-source combustion response data sequence is then segmented into time windows to generate multiple combustion cycle time window sequences. The cylinder pressure signal is scanned within each combustion cycle time window to identify the location of the local maximum value of the cylinder pressure signal, and the time point corresponding to the location is determined as the cylinder pressure peak event, generating a cylinder pressure peak event sequence. Within the same combustion cycle time window, the vibration signal of the engine body is scanned, the peak position of the vibration amplitude of the vibration signal is identified, the time point corresponding to the position is determined as the vibration response peak event, and a vibration response peak event sequence is generated. Within the combustion working cycle time window, the exhaust temperature signal is continuously detected to identify the rapid rise range of exhaust temperature. The position where the exhaust temperature change rate reaches the preset threshold is determined as the exhaust temperature rise event, and an exhaust temperature rise event sequence is generated. The fuel injection trigger signal is read within the combustion working cycle time window, and the fuel injection trigger event is identified according to the trigger time of the fuel injection trigger signal, and a fuel injection trigger event sequence is generated. A unified time index is calibrated for the cylinder pressure peak event sequence, vibration response peak event sequence, exhaust temperature rise event sequence, and fuel injection trigger event sequence. Various events are recorded according to their occurrence time to generate a combustion event time index set. The various events in the combustion event time index set are mapped to combustion propagation nodes, and the combustion propagation nodes are classified and organized according to the combustion work cycle to generate a combustion propagation node set. The generation of the combustion propagation node set specifically includes: The combustion event time index set is used to extract cylinder pressure peak events, vibration response peak events, exhaust temperature rise events, and fuel injection trigger events, and event type identifiers are established for each type of event to form a combustion event type set. Combustion propagation node numbers are established based on the various events in the combustion event type set, and the node numbers, event time indices, and event types are combined and recorded to form a combustion propagation node record set. The combustion cycle number corresponding to each combustion propagation node is extracted using the combustion propagation node record set as input, and the combustion propagation nodes are grouped according to the combustion cycle number to form a combustion propagation node group set. Within the combustion propagation node group set, the combustion propagation nodes within the same combustion cycle are arranged according to the time index order to form a combustion propagation node sequence set. The combustion propagation node sequence set is used to establish the temporal order relationship between combustion propagation nodes, and the combustion propagation node sequences are organized according to the combustion cycle to form a combustion node set.

[0020] In this embodiment, the generation of the combustion propagation path set specifically includes: Obtain the set of combustion propagation nodes, extract the time index corresponding to each combustion propagation node, and sort the combustion propagation nodes in ascending order according to the time index to generate a time-ordered sequence of combustion propagation nodes; Based on the time-ordered combustion propagation node sequence, the time difference between adjacent combustion propagation nodes in the sequence is calculated, and the node time interval information between each combustion propagation node is recorded to generate a node time interval sequence. The propagation relationship between each combustion propagation node is determined based on the node time interval sequence. When the node time interval is less than the combustion propagation time threshold, a propagation connection relationship is established between the corresponding combustion propagation nodes, and a set of propagation connection edges is generated. The connection relationships between combustion propagation nodes are organized according to the propagation connection edge set. Combustion propagation nodes connected sequentially according to the propagation connection relationship are combined to form a node propagation sequence, generating a node propagation sequence set. The propagation order of the node propagation sequence set is detected, and the node propagation sequence that satisfies the increasing relationship of node time index is determined as the combustion propagation path. Each combustion propagation path is numbered to generate a set of combustion propagation paths. The set of combustion propagation paths is categorized and organized according to the combustion cycle, and the combustion propagation path corresponding to each combustion cycle is stored to generate a set of combustion propagation paths.

[0021] In this embodiment, the generation of the combustion propagation topology specifically includes: Each combustion propagation path in the set of combustion propagation paths is parsed according to the connection order of the combustion propagation nodes, and the sequence of combustion propagation nodes contained in each combustion propagation path is extracted to generate a set of path node sequences. Extract all combustion propagation nodes from the path node sequence set, and establish a combustion propagation node set according to the node identifier. The combustion propagation node set is then determined as the topological node set in the combustion propagation topology. Based on the node connection order in the path node sequence set, extract the node connection relationship between combustion propagation nodes, record the connection relationship between adjacent combustion propagation nodes as propagation connection edges, and generate a propagation connection edge set. Based on the set of combustion propagation nodes and the set of propagation connection edges, a node connection relationship structure is established, and the combustion propagation nodes are used as topological nodes and the propagation connection edges are used as topological edges to generate the initial combustion propagation topology structure. The frequency of each propagation connection edge in the set of combustion propagation paths is counted, and the connection frequency information corresponding to each propagation connection edge is recorded according to the frequency of occurrence, thus generating a set of propagation connection edge frequencies. Write the frequency set of propagation connection edges into the corresponding propagation connection edges in the initial combustion propagation topology to generate a combustion propagation topology containing topology nodes, propagation connection edges, and propagation connection edge frequency information.

[0022] In this embodiment, the generation of the combustion propagation structure stability domain specifically includes: Read the combustion propagation topology corresponding to multiple consecutive combustion working cycles, and establish a combustion propagation topology sequence according to the order of the combustion working cycles; Based on the combustion propagation topology sequence, the propagation connection edges in each combustion propagation topology are extracted, and a unified identifier is established for the propagation connection edges in different combustion work cycles to generate a set of propagation connection edges; Based on the propagation connection edge set, record the number of times each propagation connection edge appears in the combustion propagation topology sequence, and generate the propagation connection edge occurrence count set; The frequency of each propagation connection edge in the combustion working cycle sequence is calculated from the set of occurrence counts of propagation connection edges, and a set of propagation connection edge frequencies is generated. Extract the propagation time difference corresponding to each propagation connection edge in the combustion propagation topology sequence, and record the propagation time difference data of each propagation connection edge in different combustion working cycles according to the combustion working cycle order to generate a set of propagation time difference sequences; Based on the set of propagation time difference sequences, the degree of dispersion of the propagation time difference of each propagation connection edge in the combustion working cycle sequence is calculated, and a set of propagation time difference stability is generated. The generation of the propagation time difference stability set specifically includes: The propagation time difference data of each propagation connection edge in different combustion working cycles is extracted using the propagation time difference sequence set. The propagation time difference data is then categorized and recorded according to the propagation connection edge number to generate a propagation time difference data set. The propagation time difference data corresponding to the same propagation connection edge in the propagation time difference data set are summed and averaged to generate a propagation time difference average value set. Using the propagation time difference average value set as a benchmark, the difference between the propagation time difference in the propagation time difference data set and the corresponding propagation time difference average value is calculated to generate a propagation time difference deviation set. The propagation time difference deviation set is used to determine the degree of dispersion of the propagation time difference of each propagation connection edge in the combustion working cycle sequence, and the degree of dispersion is uniformly recorded to generate a propagation time difference stability set. Based on the set of propagation connection edge frequencies and the set of propagation time difference stability, the propagation connection edges that satisfy the occurrence frequency condition and the propagation time difference stability condition are determined, and the propagation connection edges that satisfy the conditions and their corresponding combustion propagation node connection relationships are written into the combustion propagation structure stability domain to generate the combustion propagation structure stability domain. The generation of the combustion propagation structure stability domain specifically includes: The frequency data of each propagation connection edge in the combustion working cycle sequence is extracted from the propagation connection edge frequency set, and a correspondence between the propagation connection edges and their frequencies is established according to the propagation connection edge numbers, generating a propagation connection edge frequency data set. The propagation time difference stability data corresponding to each propagation connection edge is extracted from the propagation time difference stability set, and a correspondence between the propagation connection edges and their times of stability is established according to the propagation connection edge numbers, generating a propagation time difference stability data set. The propagation connection edge frequency data set and the propagation time difference stability data set are matched according to the propagation connection edge numbers, and a relationship is established between the propagation connection edges, their frequencies, and their times of stability. The corresponding relationships between the nodes are used to generate a stable data set of propagation connection edges. Propagation connection edges whose frequency of occurrence satisfies the frequency of occurrence condition and whose propagation time difference stability satisfies the propagation time difference stability condition are extracted from the stable data set to generate a stable propagation connection edge set. Based on the stable propagation connection edge set, the combustion propagation node connection relationships corresponding to each propagation connection edge are extracted, and a correspondence between the propagation connection edges and the combustion propagation node connection relationships is established to generate a stable combustion propagation connection relationship set. This stable combustion propagation connection relationship set is written into the stable domain of the combustion propagation structure, and the combustion propagation topology is constructed according to the connection relationships between the combustion propagation nodes to generate the stable domain of the combustion propagation structure.

[0023] In this embodiment, the generation of the topology deviation structure specifically includes: Obtain the combustion propagation topology and the stability domain of the combustion propagation structure corresponding to the current combustion cycle, and extract the propagation connection edges from the current combustion propagation topology to generate the current propagation connection edge set; Extract stable propagation connection edges from the stable domain of the combustion propagation structure to generate a set of stable propagation connection edges; Establish a correspondence between the current set of propagation connection edges and the stable set of propagation connection edges, and generate a set of propagation connection edge matching results. Based on the set of propagation connection edge matching results, identify propagation connection edges that exist in the current set of propagation connection edges but do not exist in the set of stable propagation connection edges, and generate a new set of propagation connection edges; Identify propagation connection edges that exist in the stable propagation connection edge set but do not exist in the current propagation connection edge set from the propagation connection edge matching result set, and generate a missing propagation connection edge set; The degree of structural deviation of the current combustion propagation topology from the stable domain of the combustion propagation structure is determined by the set of newly added propagation connection edges and the set of missing propagation connection edges, and a topology deviation determination quantity is generated. The generation of the topology deviation determination quantity specifically includes: Extract the combustion propagation node connection relationships corresponding to each newly added propagation connection edge from the newly added propagation connection edge set, and number and identify each newly added propagation connection edge to generate a new propagation connection edge data set. Extract the combustion propagation node connection relationships corresponding to each missing propagation connection edge from the missing propagation connection edge set, and number and identify each missing propagation connection edge to generate a missing propagation connection edge data set. Merge the new propagation connection edge data set and the missing propagation connection edge data set to establish a correspondence between the propagation connection edge numbers and the combustion propagation node connection relationships, generating a propagation connection edge deviation data set. Count the number of newly added propagation connection edges and the number of missing propagation connection edges in the propagation connection edge deviation data set, and record the combustion propagation node connection relationships corresponding to each propagation connection edge to generate a propagation connection edge deviation statistics set. Determine the combined deviation value of the number of newly added propagation connection edges and the number of missing propagation connection edges based on the propagation connection edge deviation statistics set, and establish a correspondence between the combined deviation value and the combustion propagation node connection relationships to generate a topology deviation judgment quantity. Based on the newly added propagation connection edge set, the missing propagation connection edge set, and the topology deviation judgment quantity, the corresponding combustion propagation node connection relationship is extracted to generate the topology deviation structure; The generation of topological deviation structures specifically includes: Extract the combustion propagation node connection relationships corresponding to each newly added propagation connection edge from the newly added propagation connection edge set, and number and identify each newly added propagation connection edge to generate a set of newly added propagation node connection relationships; extract the combustion propagation node connection relationships corresponding to each missing propagation connection edge from the missing propagation connection edge set, and number and identify each missing propagation connection edge to generate a set of missing propagation node connection relationships; merge the set of newly added propagation node connection relationships and the set of missing propagation node connection relationships, and establish a correspondence between the propagation connection edge numbers and the combustion propagation node connection relationships to generate a set of propagation node deviation connection relationships; record the combustion propagation node connection relationships and their connection directions corresponding to each propagation connection edge according to the set of propagation node deviation connection relationships to generate a set of propagation node deviation connection records; associate the set of propagation node deviation connection records with the topology deviation judgment quantity, and establish a correspondence between the topology deviation judgment quantity and the combustion propagation node connection relationships to generate a topology deviation structure.

[0024] In this embodiment, the generation of the digital health diagnosis results for the diesel engine specifically includes: Extract the deviation propagation connection edges and the corresponding combustion propagation node connection relationships from the topological deviation structure to generate a set of deviation propagation connection edges; The number of deviation propagation connections generated in the current combustion cycle is recorded based on the set of deviation propagation connections, and the number of deviation propagation connections in each combustion cycle is recorded in the order of the combustion cycle to generate a sequence of deviation propagation connection counts. The total number of propagation connection edges in the current combustion propagation topology is extracted based on the sequence of deviation propagation connection edge counts, and the topology deviation intensity result is established based on the deviation propagation connection edge counts and the total number of propagation connection edges. Extract the combustion propagation nodes that cause structural deviations from the topological deviation structure, and record the number of times each combustion propagation node appears in the deviation propagation connection edge to generate a set of the occurrence counts of deviation propagation nodes; The combustion propagation node anomaly result is established from the set of occurrences of deviation propagation nodes, and a node propagation anomaly set is generated; An indicator for determining combustion propagation anomalies was established based on the topological deviation intensity results and the set of node propagation anomalies. The current operating status of the diesel engine is generated using the results of the combustion propagation anomaly judgment index, and the digital health diagnosis results of the diesel engine are output.

[0025] Example 1: To verify the feasibility of this invention in practice, it was applied to the monitoring of the operating status of a medium-speed diesel engine in a coastal power equipment maintenance scenario. This diesel engine is responsible for the propulsion power of the ship for a long time. During continuous operation, the combustion state is affected by factors such as fuel injection state, cylinder sealing state, and intake and exhaust conditions. Therefore, the propagation relationship in the combustion process changes with the operating state. Traditional monitoring methods usually only use vibration signals or cylinder pressure signals for single feature judgment. When the combustion propagation relationship undergoes structural changes, it is often difficult to identify abnormal propagation paths in time, which can easily lead to insufficient identification of combustion anomalies. In this operating scenario, by arranging cylinder pressure sensing devices, vibration detection devices, crankshaft speed detection devices, exhaust temperature detection devices, and fuel injection trigger signal acquisition devices at corresponding positions on the diesel engine body, various combustion response information generated during the diesel engine operation is continuously collected, thereby obtaining cylinder pressure signals, engine vibration signals, crankshaft speed signals, exhaust temperature signals, and fuel injection trigger signals. After uniform processing of these signals, a multi-source combustion response data sequence is generated.

[0026] In actual operation, the collected multi-source combustion response data is first preprocessed to ensure that the signals from different detection devices are consistent on the time axis. At the same time, environmental interference in the signals is filtered out so that the data can truly reflect the dynamic changes in the combustion process. The preprocessed multi-source combustion response data sequence is divided into time windows according to the diesel engine combustion working cycle. The start and end positions of each combustion working cycle are determined by the change of crankshaft speed signal, so that each time window can accurately correspond to a complete combustion process. Within each combustion working cycle, the cylinder pressure signal is scanned to identify the peak position in the pressure change process, and the corresponding time point is identified as the cylinder pressure peak event. At the same time, the engine vibration signal is scanned to identify the peak position in the vibration amplitude change, and the corresponding time point is identified as the vibration response peak event. Meanwhile, the change process of the exhaust temperature signal is continuously detected. When the exhaust temperature rises rapidly, the exhaust temperature rise event is identified. Combined with the trigger time of the fuel injection trigger signal, the fuel injection trigger event is identified. By uniformly indexing and calibrating these combustion response events, different types of combustion events can be mapped to combustion propagation nodes, thus forming a set of combustion propagation nodes.

[0027] After obtaining the set of combustion propagation nodes, propagation connections between nodes are established based on the time index relationship between each node. When the time interval between adjacent nodes meets the combustion propagation time condition, a propagation connection edge is established between the corresponding nodes, and multiple nodes are combined according to the propagation order to form a combustion propagation path. As multiple combustion work cycles continue, different combustion propagation paths are continuously recorded, thus forming a set of combustion propagation paths. Based on this, combustion propagation nodes are used as topological nodes, and propagation connection edges between nodes are used as topological edges. After parsing the set of combustion propagation paths, a combustion propagation topology structure is constructed, so that the propagation relationship in the diesel engine combustion process can be expressed in a structured form. In this way, the propagation order and propagation relationship between different combustion response events can be clearly represented in the form of a topological structure, thus reflecting the propagation characteristics in the combustion process more intuitively.

[0028] As the diesel engine continues to run, the occurrence of each propagation connection edge is statistically analyzed across multiple combustion cycles, and the changes in the propagation time difference between these edges are recorded. By analyzing the frequency of occurrence of the propagation connection edges and the stability of the propagation time difference, a stable domain of the combustion propagation structure reflecting a stable combustion state can be gradually formed. When the diesel engine is in a normal combustion state, most propagation connection edges will continuously appear across multiple combustion cycles, and the propagation time difference will remain relatively stable. The stable domain of the combustion propagation structure constructed under these conditions can accurately reflect the combustion propagation structure characteristics of the diesel engine during stable operation. When the combustion state changes, such as changes in fuel injection atomization or cylinder sealing, some propagation connection edges will change, resulting in structural differences between the current combustion propagation topology and the stable domain.

[0029] During operation monitoring, the combustion propagation topology formed by the current combustion cycle is structurally matched with the stable domain of the combustion propagation structure. By comparing the differences between the current set of propagation connection edges and the set of stable propagation connection edges, newly added and missing propagation connection edges can be identified, and a topology deviation judgment quantity can be further formed. When a newly added propagation connection edge appears, it indicates that a new propagation path has been generated in the combustion propagation process; when a stable propagation connection edge is missing, it indicates that the original combustion propagation relationship has been interrupted. By simultaneously analyzing newly added and missing propagation connection edges, the overall change of the combustion propagation structure can be characterized, and a topology deviation structure can be constructed accordingly. This topology deviation structure can directly reflect the abnormal connection relationships between combustion propagation nodes, so that combustion propagation anomalies can be expressed in structural form.

[0030] During continuous monitoring, by recording the changes in the combustion propagation structure over different time periods, it can be observed that the topological deviation structure of the diesel engine changes little under stable operating conditions. However, during operating phases where the combustion state changes, the number of newly added and missing propagation connection edges gradually increases, and the topological deviation judgment value changes significantly accordingly. By further extracting the deviation propagation connection edges and corresponding combustion propagation nodes from the topological deviation structure, the location of anomalies in the combustion propagation relationship can be determined. Combined with the changing trend of the deviation propagation connection edges in the continuous combustion working cycle, the operating state of the diesel engine can be judged. In actual operation monitoring, by recording the topological deviation structure formed during continuous operating time periods, it can be found that when the diesel engine combustion process remains stable, the propagation relationship between each combustion propagation node remains consistent over a long period of time. However, when the combustion state is abnormal, the propagation connection relationship changes significantly in a short period of time, thus reflecting the abnormal development trend of the combustion state in a timely manner.

[0031] As can be seen from the above application process, this invention constructs combustion propagation nodes, combustion propagation paths, and combustion propagation topology, and performs structural matching of the current combustion propagation relationship in conjunction with the combustion propagation structure stability domain. This enables a structured analysis of changes in the combustion propagation relationship of diesel engines. Compared with traditional methods that rely on single signal features or simple statistical indicators, this method can identify abnormal changes from the perspective of the overall structure of the combustion propagation relationship. This allows the diagnosis of diesel engine operating status to more comprehensively reflect the changes in propagation characteristics during the combustion process, thereby providing more reliable technical support for diesel engine operating status monitoring and maintenance.

[0032] Table 1. Performance Comparison of Diesel Engine Combustion Anomaly Diagnosis Methods

[0033] As shown in Table 1, the traditional vibration feature threshold diagnostic method has an accuracy of 74.6% in identifying diesel engine combustion anomalies. This method mainly relies on the characteristics of engine vibration signals for judgment and can reflect some changes in mechanical state. However, it does not make sufficient use of the correlation of multi-source signals in the combustion propagation process. The multi-signal statistical feature diagnostic method improves the accuracy to 81.3% by introducing multiple signals such as cylinder pressure, vibration, and temperature for joint statistical analysis. In contrast, this invention improves the anomaly identification accuracy to 87.2% by constructing combustion propagation nodes, combustion propagation paths, and combustion propagation topology to structurally express the propagation relationship between multi-source combustion response events. This improvement mainly comes from the introduction of combustion propagation structure information, which allows the time propagation relationship between combustion events to be analyzed as a whole, thereby reducing the interference of single signal fluctuations on the diagnostic results.

[0034] The method of this invention also demonstrates significant advantages in terms of false alarm rate and false negative rate. The false alarm rate of the vibration feature threshold method reaches 14.8%, mainly because vibration signals are easily affected by factors such as mechanical resonance and load changes, leading to unstable anomaly judgment. The multi-signal statistical feature method reduces single-signal interference by introducing multiple signals, but still has a false alarm rate of 11.2% under complex operating conditions. In contrast, this invention constructs a stable domain of combustion propagation structure in a continuous combustion working cycle and uses the stable domain to perform structural matching on the current combustion propagation topology, making anomaly judgment based on deviation of propagation structure rather than single signal change. Therefore, the false alarm rate is reduced to 7.6%, and the false negative rate is also reduced to 5.2%. This shows that the method can more stably identify the propagation relationship of combustion anomalies.

[0035] Regarding diagnostic time, the average single-cycle diagnostic time of the method of this invention is 108 milliseconds, which is slightly increased compared to traditional methods. This is because the present invention requires extraction of combustion propagation nodes, construction of propagation paths, and analysis of topology during the diagnostic process. However, this calculation process still remains within the allowable range of real-time monitoring and does not significantly delay diesel engine operation monitoring. In contrast, the present invention performs better in terms of early identification of combustion anomalies, being able to identify potential anomalies on average 7.2 combustion cycles in advance, while traditional vibration threshold methods can only detect abnormal changes about 4 combustion cycles in advance. This is because changes in the combustion propagation structure often precede changes in macroscopic signals such as vibration and temperature. Therefore, analyzing the propagation topology can detect abnormal propagation paths in the combustion process earlier.

[0036] Regarding the ability to identify abnormal locations, the method of this invention achieves an abnormal propagation location identification rate of 73.9%, which is significantly higher than that of traditional methods. Traditional vibration threshold methods can only detect overall vibration changes and are difficult to locate specific combustion propagation nodes, with an abnormal propagation location identification rate of only 46.5%. Although multi-signal statistical methods can infer the source of anomalies through feature changes, they still lack a clear description of propagation relationships, with an identification rate of 58.4%. This invention directly records newly added and missing propagation connection edges through topological deviation structures and combines them with the connection relationships between combustion propagation nodes, enabling abnormal propagation paths to be directly identified in the topological structure. Therefore, it can more accurately locate abnormal propagation nodes.

[0037] Furthermore, the method of this invention also has significant advantages in identifying multi-source signals by utilizing dimensional and structural propagation relationships. Traditional vibration thresholding methods only use a single vibration signal for judgment, while multi-signal statistical methods typically only perform joint analysis on three types of signals. This invention simultaneously integrates cylinder pressure, engine vibration, crankshaft speed, exhaust temperature, and fuel injection trigger signals, mapping these signals as combustion propagation nodes. By constructing a combustion propagation topology through propagation connections, it can identify changes in structural relationships during combustion propagation. The data in the table shows that the recognition rate of structural propagation relationships by this invention reaches 68.7%, far exceeding the 21.6% of traditional statistical methods, indicating that this method has a stronger ability to express the combustion propagation mechanism.

[0038] The comprehensive data shows that this invention constructs combustion propagation nodes, combustion propagation paths, and combustion propagation topology, and uses the combustion propagation structure stability domain and topology deviation judgment quantity to perform structural matching of the current combustion state. This enables the effective identification of abnormal propagation relationships in the diesel engine combustion process. This method not only improves the accuracy of diesel engine anomaly diagnosis but also enhances the ability to identify abnormal propagation paths. It enables diesel engine operating status monitoring to more comprehensively reflect the structural change characteristics in the combustion process, thus verifying the effectiveness and technical advantages of this invention in practical applications.

[0039] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A digital health diagnosis method for diesel engines based on multi-source data fusion, characterized in that, Includes the following steps: Collect cylinder pressure signals, engine vibration signals, crankshaft speed signals, exhaust temperature signals, and fuel injection trigger signals generated during diesel engine operation, and preprocess them to generate multi-source combustion response data sequences; Based on the diesel engine combustion cycle, the multi-source combustion response data sequence is divided into time windows. Within each combustion cycle, cylinder pressure peak events, vibration response peak events, exhaust temperature rise events, and fuel injection trigger events are identified. These events are mapped to combustion propagation nodes, generating a set of combustion propagation nodes. Based on the time index relationship of each node in the combustion propagation node set, establish the propagation connection relationship between the nodes and generate a set of combustion propagation paths; Construct a combustion propagation topology based on the set of combustion propagation paths; Statistical analysis of the combustion propagation topology is performed over multiple consecutive combustion cycles to calculate the frequency of combustion propagation path occurrence and the stability of propagation time difference, thereby generating a stable domain for the combustion propagation structure. The current combustion propagation topology is matched with the stability domain of the combustion propagation structure, the topology deviation determination quantity is calculated, and the topology deviation structure is generated. The health status of the diesel engine is determined based on the topological deviation structure, and a digital health diagnosis result for the diesel engine is generated.

2. The method for digital health diagnosis of diesel engines based on multi-source data fusion according to claim 1, characterized in that, The preprocessing specifically includes data synchronization calibration, noise filtering, outlier removal, data resampling, signal normalization, and time series alignment.

3. The method for digital health diagnosis of diesel engines based on multi-source data fusion according to claim 1, characterized in that, The generation of the combustion propagation node set specifically includes: The multi-source combustion response data sequence is divided into cycles according to the crankshaft speed signal. The start and end positions of each combustion cycle are determined based on the crankshaft speed change. The multi-source combustion response data sequence is then segmented into time windows to generate multiple combustion cycle time window sequences. The cylinder pressure signal is scanned within each combustion cycle time window to identify the location of the local maximum value of the cylinder pressure signal, and the time point corresponding to the location is determined as the cylinder pressure peak event, generating a cylinder pressure peak event sequence. Within the same combustion cycle time window, the vibration signal of the engine body is scanned, the peak position of the vibration amplitude of the vibration signal is identified, the time point corresponding to the position is determined as the vibration response peak event, and a vibration response peak event sequence is generated. Within the combustion working cycle time window, the exhaust temperature signal is continuously detected to identify the rapid rise range of exhaust temperature. The position where the exhaust temperature change rate reaches the preset threshold is determined as the exhaust temperature rise event, and an exhaust temperature rise event sequence is generated. The fuel injection trigger signal is read within the combustion working cycle time window, and the fuel injection trigger event is identified according to the trigger time of the fuel injection trigger signal, and a fuel injection trigger event sequence is generated. A unified time index is calibrated for the cylinder pressure peak event sequence, vibration response peak event sequence, exhaust temperature rise event sequence, and fuel injection trigger event sequence. Various events are recorded according to their occurrence time to generate a combustion event time index set. The various events in the combustion event time index set are mapped to combustion propagation nodes, and the combustion propagation nodes are categorized and organized according to the combustion work cycle to generate a combustion propagation node set.

4. The method for digital health diagnosis of diesel engines based on multi-source data fusion according to claim 1, characterized in that, The generation of the set of combustion propagation paths specifically includes: Obtain the set of combustion propagation nodes, extract the time index corresponding to each combustion propagation node, and sort the combustion propagation nodes in ascending order according to the time index to generate a time-ordered sequence of combustion propagation nodes; Based on the time-ordered combustion propagation node sequence, the time difference between adjacent combustion propagation nodes in the sequence is calculated, and the node time interval information between each combustion propagation node is recorded to generate a node time interval sequence. The propagation relationship between each combustion propagation node is determined based on the node time interval sequence. When the node time interval is less than the combustion propagation time threshold, a propagation connection relationship is established between the corresponding combustion propagation nodes, and a set of propagation connection edges is generated. The connection relationships between combustion propagation nodes are organized according to the propagation connection edge set. Combustion propagation nodes connected sequentially according to the propagation connection relationship are combined to form a node propagation sequence, generating a node propagation sequence set. The propagation order of the node propagation sequence set is detected, and the node propagation sequence that satisfies the increasing relationship of node time index is determined as the combustion propagation path. Each combustion propagation path is numbered to generate a set of combustion propagation paths. The set of combustion propagation paths is categorized and organized according to the combustion cycle, and the combustion propagation path corresponding to each combustion cycle is stored to generate a set of combustion propagation paths.

5. The method for digital health diagnosis of diesel engines based on multi-source data fusion according to claim 1, characterized in that, The generation of the combustion propagation topology specifically includes: Each combustion propagation path in the set of combustion propagation paths is parsed according to the connection order of the combustion propagation nodes, and the sequence of combustion propagation nodes contained in each combustion propagation path is extracted to generate a set of path node sequences. Extract all combustion propagation nodes from the path node sequence set, and establish a combustion propagation node set according to the node identifier. The combustion propagation node set is then determined as the topological node set in the combustion propagation topology. Based on the node connection order in the path node sequence set, extract the node connection relationship between combustion propagation nodes, record the connection relationship between adjacent combustion propagation nodes as propagation connection edges, and generate a propagation connection edge set. Based on the set of combustion propagation nodes and the set of propagation connection edges, a node connection relationship structure is established, and the combustion propagation nodes are used as topological nodes and the propagation connection edges are used as topological edges to generate the initial combustion propagation topology structure. The frequency of each propagation connection edge in the set of combustion propagation paths is counted, and the connection frequency information corresponding to each propagation connection edge is recorded according to the frequency of occurrence, thus generating a set of propagation connection edge frequencies. Write the frequency set of propagation connection edges into the corresponding propagation connection edges in the initial combustion propagation topology to generate a combustion propagation topology containing topology nodes, propagation connection edges, and propagation connection edge frequency information.

6. The method for digital health diagnosis of diesel engines based on multi-source data fusion according to claim 1, characterized in that, The generation of the combustion propagation structure stability domain specifically includes: Read the combustion propagation topology corresponding to multiple consecutive combustion working cycles, and establish a combustion propagation topology sequence according to the order of the combustion working cycles; Based on the combustion propagation topology sequence, the propagation connection edges in each combustion propagation topology are extracted, and a unified identifier is established for the propagation connection edges in different combustion work cycles to generate a set of propagation connection edges; Based on the propagation connection edge set, record the number of times each propagation connection edge appears in the combustion propagation topology sequence, and generate the propagation connection edge occurrence count set; The frequency of each propagation connection edge in the combustion working cycle sequence is calculated from the set of occurrence counts of propagation connection edges, and a set of propagation connection edge frequencies is generated. Extract the propagation time difference corresponding to each propagation connection edge in the combustion propagation topology sequence, and record the propagation time difference data of each propagation connection edge in different combustion working cycles according to the combustion working cycle order to generate a set of propagation time difference sequences; Based on the set of propagation time difference sequences, the degree of dispersion of the propagation time difference of each propagation connection edge in the combustion working cycle sequence is calculated, and a set of propagation time difference stability is generated. Based on the set of propagation connection edge frequencies and the set of propagation time difference stability, the propagation connection edges that satisfy the occurrence frequency condition and the propagation time difference stability condition are determined, and the propagation connection edges that satisfy the conditions and their corresponding combustion propagation node connection relationships are written into the combustion propagation structure stability domain to generate the combustion propagation structure stability domain.

7. The method for digital health diagnosis of diesel engines based on multi-source data fusion according to claim 1, characterized in that, The generation of the topology deviation structure specifically includes: Obtain the combustion propagation topology and the stability domain of the combustion propagation structure corresponding to the current combustion cycle, and extract the propagation connection edges from the current combustion propagation topology to generate the current propagation connection edge set; Extract stable propagation connection edges from the stable domain of the combustion propagation structure to generate a set of stable propagation connection edges; Establish a correspondence between the current set of propagation connection edges and the stable set of propagation connection edges, and generate a set of propagation connection edge matching results. Based on the set of propagation connection edge matching results, identify propagation connection edges that exist in the current set of propagation connection edges but do not exist in the set of stable propagation connection edges, and generate a new set of propagation connection edges; Identify propagation connection edges that exist in the stable propagation connection edge set but do not exist in the current propagation connection edge set from the propagation connection edge matching result set, and generate a missing propagation connection edge set; The degree of structural deviation of the current combustion propagation topology from the stable domain of the combustion propagation structure is determined by the set of newly added propagation connection edges and the set of missing propagation connection edges, and a topology deviation determination quantity is generated. Based on the newly added propagation connection edge set, the missing propagation connection edge set, and the topology deviation judgment quantity, the corresponding combustion propagation node connection relationship is extracted to generate the topology deviation structure.

8. The method for digital health diagnosis of diesel engines based on multi-source data fusion according to claim 1, characterized in that, The generation of the digital health diagnosis results for the diesel engine specifically includes: Extract the deviation propagation connection edges and the corresponding combustion propagation node connection relationships from the topological deviation structure to generate a set of deviation propagation connection edges; The number of deviation propagation connections generated in the current combustion cycle is recorded based on the set of deviation propagation connections, and the number of deviation propagation connections in each combustion cycle is recorded in the order of the combustion cycle to generate a sequence of deviation propagation connection counts. The total number of propagation connection edges in the current combustion propagation topology is extracted based on the sequence of deviation propagation connection edge counts, and the topology deviation intensity result is established based on the deviation propagation connection edge counts and the total number of propagation connection edges. Extract the combustion propagation nodes that cause structural deviations from the topological deviation structure, and record the number of times each combustion propagation node appears in the deviation propagation connection edge to generate a set of the occurrence counts of deviation propagation nodes; The combustion propagation node anomaly result is established from the set of occurrences of deviation propagation nodes, and a node propagation anomaly set is generated; An indicator for determining combustion propagation anomalies was established based on the topological deviation intensity results and the set of node propagation anomalies. The current operating status of the diesel engine is generated using the results of the combustion propagation anomaly judgment index, and the digital health diagnosis results of the diesel engine are output.

Images