Crane abnormity monitoring method and system and application thereof

Abstract

The invention discloses a crane abnormity monitoring method and system and application thereof.According to the scheme, environment monitoring data and parameter monitoring data and state monitoring data related to a crane are fused, and when the crane is abnormal, the data are judged, so that the crane abnormity is monitored; the method comprises the following steps of: roughly judging whether a factor which makes the state of the crane abnormal is an internal factor of the crane or is caused by sudden change of an environmental factor, providing abnormity checking assistance to a certain extent for background management personnel or field operation personnel, and providing assistance for providing a targeted abnormity checking direction for maintenance personnel; besides, according to the scheme, after the database containing the abnormal data set is established, when the crane work is monitored subsequently, the environment monitoring data and the work monitoring data which are obtained in real time can be matched with the abnormal data set, and a pre-prompt is provided when the crane is possibly abnormal subsequently; and a reference is provided for risk avoidance during operation of an operator.

Description

Technical Field

[0001] The present invention relates to the fields of crane monitoring technology and crane control technology, and in particular to a crane abnormality monitoring method, system and application thereof. Background Art

[0002] As an important transportation equipment, cranes are widely used in workshops, warehouses, ports, construction sites and other scenes. In terms of application sites, cranes can be deployed not only indoors, but also outdoors. Compared with indoors, cranes are more affected by more complicated working factors when working outdoors, such as environmental wind, temperature, humidity, rain, etc. These environmental factors often have characteristics such as variable and nonlinear changes. Therefore, when cranes work outdoors, their safety monitoring and abnormal monitoring need to be more precise and strict. Most of the literature disclosed so far only monitors the current, vibration or noise of the crane when it is working, ignoring whether the possible causes of such problems are external or internal, which makes it difficult to timely warn and monitor the potential dangers that may occur when the crane is working outdoors. Especially in the reuse of historical data, there are few literatures disclosed at present. Therefore, how to optimize the abnormal monitoring of the crane so that when an abnormality occurs, abnormal problems, parameters and possible potential factors can be fed back in time is a research topic with positive and practical significance. Summary of the invention

[0003] In view of this, the purpose of the present invention is to propose a crane abnormality monitoring method, system and application thereof, which are reliable in implementation, efficient in response, have good reliability of feedback information and can provide early warning of abnormal conditions based on historical data reuse.

[0004] In order to achieve the above technical objectives, the technical solution adopted by the present invention is:

[0005] A method for monitoring abnormalities of a crane, wherein the crane is deployed in an outdoor work site, an environmental monitoring module is deployed in the work site, and a monitoring sensor group is deployed on a characteristic working part of the crane, comprising:

[0006] S01, the environmental monitoring module monitors the environmental parameters in the workplace in real time and generates environmental monitoring data;

[0007] S02, responding to the work start signal of the crane, recording the working parameters of the crane and generating parameter monitoring data; at the same time, monitoring the state of the crane through the monitoring sensor group deployed on the crane and generating state monitoring data;

[0008] S03, acquiring the status monitoring data in real time, judging the status monitoring data according to preset conditions, and generating a judgment result to identify whether the crane is working abnormally. When the crane is working abnormally, a multimodal fusion model based on the environmental monitoring data, the working monitoring data and the status monitoring data is constructed, and then the parameter monitoring data, the status monitoring data and the environmental monitoring data are comprehensively analyzed according to preset conditions to generate an analysis result to distinguish whether the working state change caused by the abnormal operation of the crane is induced by the internal factors of the crane and / or by the environmental factors;

[0009] S04. Obtain the judgment result. When it points to abnormal operation of the crane, collect and record the environmental monitoring data, parameter monitoring data and status monitoring data of the preset time period before and after the abnormality occurs, as well as the analysis results, and then attach a timestamp and mark as abnormal information to form an abnormal data group, which is stored in the database.

[0010] As a possible implementation mode, further, the environmental monitoring module described in this solution includes one or more of a wind speed sensor, a humidity sensor, a temperature sensor, and a rain gauge sensor.

[0011] As a possible implementation, further, in the present solution S01, the environmental monitoring data collected and generated by the environmental monitoring module is recorded in the form of a data set and marked with a relevant timestamp, and the record is represented as follows:

[0012] E(t)=[E1(t), E2(t),…,E n (t)]

[0013] Among them, t is the time point when the environmental monitoring module collects data, E1(t), E2(t), ..., E n (t) is the data item collected by the sensor of the environment monitoring module, which includes data collected by one or more sensors selected from the group consisting of a wind speed sensor, a humidity sensor, a temperature sensor, and a rain gauge sensor.

[0014] As a preferred implementation option, preferably, the environmental monitoring data described in this scheme is also smoothed by a sliding average method to partially eliminate noise and abnormal data. The model is defined as follows:

[0015]

[0016] is the environmental data after sliding average processing, W is the sliding window size, which is used to represent the number of historical data points involved in the calculation; E i (j) is the collected value of the i-th environmental data at time j.

[0017] As a better implementation option, preferably, the crane described in this scheme is a bridge crane, and the characteristic working parts include a gantry, a luffing trolley, a hook drive motor, and a translation drive motor. The hook drive motor is deployed on the luffing trolley, and the translation drive motor is connected to both sides of the lower part of the gantry and is used to drive the gantry to translate on the track of the work site.

[0018] As a preferred implementation option, preferably, the monitoring sensor group described in this scheme includes: a vibration sensor and / or a sound sensor, wherein the vibration sensor is used to collect the vibration conditions of the characteristic working parts of the crane when it is working, and the sound sensor is used to collect the noise conditions of the characteristic working parts of the crane when it is working.

[0019] As a preferred implementation option, preferably, in this solution S02, the operating parameters of the crane include: driving current and braking torque;

[0020] The parameter monitoring data includes driving current data and braking torque data;

[0021] The condition monitoring data includes vibration data and noise data.

[0022] As a preferred implementation option, preferably, the data in the parameter monitoring data and the state monitoring data described in this solution are all attached with the timestamp when they were generated; and their records are respectively expressed as follows:

[0023] S(t)=[S1(t), S2(t ) ]

[0024] Z(t)=[Z1(t),Z2(t)]

[0025] Where, t is time, S(t) is parameter monitoring data, S 1( t ) , S 2( t ) They are driving current data and braking torque data respectively; Z ( t ) is the state monitoring data, Z1(t ) 、Z2(t ) They are vibration data and noise data respectively.

[0026] As a preferred implementation option, preferably, in the present scheme S03, a judgment result is generated by comparing the status monitoring data with the corresponding threshold value. When one of the data in the status monitoring data exceeds the preset corresponding threshold value, a judgment result pointing to an operating abnormality is generated, and the characteristic working components corresponding to the operating abnormality are marked in the judgment result. Otherwise, a judgment result pointing to normal operation is generated.

[0027] As a preferred implementation option, preferably, in this solution S03, when the judgment result points to work abnormality, a multimodal fusion model based on environmental monitoring data, work monitoring data and status monitoring data is constructed, which is set as a multimodal feature matrix F(t), which is expressed as:

[0028] F(t)=[E(t),S(t),Z(t)]

[0029] Among them, E(t) is the environmental monitoring data, S(t) is the parameter monitoring data, and Z(t) is the state monitoring data;

[0030] The parameter monitoring data, state monitoring data and environmental monitoring data in the multimodal feature matrix F(t) are comprehensively analyzed through the trained random forest algorithm or neural network model to generate analysis results to distinguish whether the working state changes caused by abnormal operation of the crane are induced by internal factors of the crane and / or by environmental factors; the internal factors are the working parameters of the crane, and the environmental factors are the environmental parameters when the crane is working.

[0031] As a preferred implementation option, preferably, in this solution S03, the parameter monitoring data, state monitoring data and environmental monitoring data in the multimodal feature matrix F(t) are comprehensively analyzed by a trained random forest algorithm, which includes:

[0032] The random forest algorithm is used to construct an anomaly classifier. The goal is to classify whether the abnormality is caused by the internal factors of the crane itself. The label Y is set and its value is assumed to be 0 or 1. 1 means the abnormality is caused by the internal factors of the crane itself, and 0 means no abnormality.

[0033] Construct K decision trees, and train each tree based on randomly sampled data from the training set (F(t), Y) until convergence;

[0034] Input the multimodal feature matrix F(t) to be detected, and let all decision trees give prediction results respectively

[0035] Aggregate the prediction results and use the voting method to output the final classification result, which is defined as follows:

[0036]

[0037] Among them, I(·) is the indicator function. When the assumption condition is met, that is, when the crane’s internal factors cause abnormal operation, its value is 1, otherwise it is 0; is the final prediction result, which is set to the analysis result, and k is the number of decision trees involved in the prediction.

[0038] The judgment of abnormal inducing factors in the present application is not limited to the random forest algorithm. It can also detect the data in the multimodal feature matrix F(t) through a trained neural network model to obtain a judgment result that the resulting working state change is induced by internal factors of the crane and / or by environmental factors.

[0039] As a preferred implementation option, preferably, in this solution S04, when the judgment result points to abnormal operation of the crane, the environmental monitoring data, parameter monitoring data and state monitoring data of the preset time period before and after the abnormality occurs are collected and recorded, and the analysis results are then added with a timestamp and marked as abnormal information to form an abnormal data group, which is defined by an abnormal feature vector A in the following form:

[0040] A k =[E k , S k , Z k , t k ,y k ]

[0041] Among them, E k , S k , Z k , t k ,y k They are the environmental monitoring data, work monitoring data, state monitoring data, timestamp when the anomaly occurs, and anomaly information label at the time of the kth work anomaly respectively; finally, the anomaly feature vector A k Stored in the database.

[0042] As a preferred implementation option, preferably, the monitoring method further includes:

[0043] S05. Acquire environmental monitoring data, work monitoring data and status monitoring data in real time, match them with abnormal data groups in the database, generate matching results, and generate risk warning information and output when the matching results meet the preset requirements.

[0044] As a preferred implementation option, preferably, in S05, the matching degree between the environmental monitoring data, the working monitoring data and the state monitoring data and the abnormal data group in the database is determined by similarity calculation, which includes:

[0045] The environmental monitoring data E current , work monitoring data S current Set the parameter F current =[E current , S current ], which corresponds to the abnormal feature vector A of the abnormal data group in the database k =[E k , Sk The similarity of ] is defined as follows:

[0046]

[0047] Among them, α and β are weight coefficients, which are used to reflect the influence weight of the environment and the working state of the crane, and ||·|| is the Euclidean norm of the vector;

[0048] When the calculated similarity Sim(F current , A k ) exceeds the preset threshold λ, a risk warning message is generated and output.

[0049] Based on the above, this solution also proposes a crane operation control method, which is applied with the above-mentioned crane abnormality monitoring method, and further includes:

[0050] An abnormal data group is obtained, and according to the environmental monitoring data and parameter monitoring data therein, thresholds of the operating parameters of the crane under the corresponding environmental monitoring data and parameter monitoring data are adjusted.

[0051] Based on the above, this solution also proposes a crane abnormality monitoring system, which includes:

[0052] The environmental monitoring module is deployed in the working area of ​​the crane and is used to collect environmental parameters and generate environmental monitoring data;

[0053] A monitoring sensor group is deployed on characteristic working parts of the crane and is used to collect vibration data and noise data when the crane is working to generate status monitoring data;

[0054] A working monitoring module is used to respond to a working start signal of the crane, record working parameters of the crane, and generate parameter monitoring data;

[0055] The data judgment module is used to obtain the status monitoring data in real time, judge the status monitoring data according to the preset conditions, and generate a judgment result to identify whether the crane is working abnormally;

[0056] The data fusion module is used to obtain the judgment result. When it points to the abnormal operation of the crane, a multimodal fusion model based on environmental monitoring data, work monitoring data and status monitoring data is constructed;

[0057] The data investigation module is used to comprehensively analyze the parameter monitoring data, the state monitoring data and the environmental monitoring data according to the preset conditions, and generate the analysis results to distinguish whether the abnormal working state of the crane is caused by the internal factors of the crane and / or by the environmental factors;

[0058] The data recording module is used to obtain the judgment result. When it points to abnormal operation of the crane, it collects and records the environmental monitoring data, parameter monitoring data and status monitoring data of the preset time period before and after the abnormality occurs, as well as the analysis results, and then attaches a timestamp and marks it as abnormal information to form an abnormal data group and store it in the database.

[0059] By adopting the above-mentioned technical scheme, compared with the prior art, the present invention has the following beneficial effects: this scheme integrates environmental monitoring data with parameter monitoring data and status monitoring data related to the crane. When an abnormality occurs in the crane, these data are judged to roughly determine whether the factor causing the abnormal state of the crane is its own internal factor or caused by a sudden change in environmental factors, thereby providing a certain degree of abnormality troubleshooting assistance for back-end management personnel or on-site operators, and providing targeted abnormality troubleshooting direction for maintenance personnel; in addition, this scheme also establishes a database containing an abnormal data group, and when the crane work is subsequently monitored, the real-time environmental monitoring data and work monitoring data can be matched with the abnormal data group, so as to provide advance prompts when the crane may have abnormalities in the future, and provide a reference for operators to avoid risks during operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0061] Figure 1 It is a brief implementation flow diagram of the abnormal monitoring method of this scheme;

[0062] Figure 2 It is a brief implementation flow diagram of the abnormal monitoring method of this scheme when it is applied to the crane work control method;

[0063] Figure 3 This is a schematic diagram of the unit module connection of the abnormal monitoring system of this scheme. DETAILED DESCRIPTION

[0064] The present invention will be further described in detail below in conjunction with the accompanying drawings and examples. It is particularly noted that the following examples are only used to illustrate the present invention, but are not intended to limit the scope of the present invention. Similarly, the following examples are only partial embodiments of the present invention rather than all embodiments, and all other embodiments obtained by those of ordinary skill in the art without creative work are within the scope of protection of the present invention.

[0065] like Figure 1 As shown, this embodiment provides a crane abnormality monitoring method, wherein the crane is deployed in an outdoor work site, an environmental monitoring module is deployed in the work site, and a monitoring sensor group is deployed on a characteristic working part of the crane, which includes:

[0066] S01, the environmental monitoring module monitors the environmental parameters in the workplace in real time and generates environmental monitoring data;

[0067] S02, responding to the work start signal of the crane, recording the working parameters of the crane and generating parameter monitoring data; at the same time, monitoring the state of the crane through the monitoring sensor group deployed on the crane and generating state monitoring data;

[0068] S03, acquiring the status monitoring data in real time, judging the status monitoring data according to preset conditions, and generating a judgment result to identify whether the crane is working abnormally. When the crane is working abnormally, a multimodal fusion model based on the environmental monitoring data, the working monitoring data and the status monitoring data is constructed, and then the parameter monitoring data, the status monitoring data and the environmental monitoring data are comprehensively analyzed according to preset conditions to generate an analysis result to distinguish whether the working state change caused by the abnormal operation of the crane is induced by the internal factors of the crane and / or by the environmental factors;

[0069] S04. Obtain the judgment result. When it points to abnormal operation of the crane, collect and record the environmental monitoring data, parameter monitoring data and status monitoring data of the preset time period before and after the abnormality occurs, as well as the analysis results, and then attach a timestamp and mark as abnormal information to form an abnormal data group, which is stored in the database.

[0070] This solution integrates environmental monitoring data with crane-related parameter monitoring data and status monitoring data. When an abnormality occurs in the crane, this solution can judge these data to roughly determine whether the abnormal factor causing the crane status is its own internal factors or due to sudden changes in environmental factors. This solution can provide a certain degree of abnormality troubleshooting assistance to back-end managers or on-site operators, and provide targeted abnormality troubleshooting directions for maintenance personnel.

[0071] As an example of environmental monitoring and as a possible implementation method, further, the environmental monitoring module described in this solution includes one or more of a wind speed sensor, a humidity sensor, a temperature sensor, and a rain gauge sensor.

[0072] By monitoring environmental parameters, it helps to improve the comprehensiveness of factors monitored during crane operation monitoring and provide better and broader data support for subsequent abnormal analysis of the crane.

[0073] In terms of data integration, as a possible implementation method, further, in the present solution S01, the environmental monitoring data collected and generated by the environmental monitoring module is recorded in the form of a data set and marked with a relevant timestamp, and the record is represented as follows:

[0074] E(t)=[E1(t), E2(t),…,E n (t)]

[0075] Among them, t is the time point when the environmental monitoring module collects data, E1(t), E2(t), ..., E n (t) is the data item collected by the sensor of the environment monitoring module, which includes data collected by one or more sensors selected from the group consisting of a wind speed sensor, a humidity sensor, a temperature sensor, and a rain gauge sensor.

[0076] In order to avoid data collection anomalies caused by some interference factors and improve the overall reliability of the data, as a better implementation option, preferably, the environmental monitoring data described in this scheme is also smoothed by a sliding average method to partially eliminate noise and abnormal data. The model is defined as follows:

[0077]

[0078] is the environmental data after sliding average processing, W is the sliding window size, which is used to represent the number of historical data points involved in the calculation; E i (j) is the collected value of the i-th environmental data at time j.

[0079] In terms of implementation objects, as a better implementation option, preferably, the crane described in this scheme is a bridge crane, and the characteristic working parts include a gantry, a luffing trolley, a hook drive motor, and a translation drive motor. The hook drive motor is deployed on the luffing trolley, and the translation drive motor is connected to both sides of the lower part of the gantry and is used to drive the gantry to translate on the track of the work site.

[0080] As for the status collection of the crane itself, as a better implementation option, preferably, the monitoring sensor group described in this scheme includes: a vibration sensor and / or a sound sensor, the vibration sensor is used to collect the vibration conditions of the characteristic working parts of the crane when it is working, and the sound sensor is used to collect the noise conditions of the characteristic working parts of the crane when it is working.

[0081] This solution monitors the vibration and noise conditions of the crane when it is working, which serve as manifestations of crane abnormalities and can serve as judgment conditions for crane abnormalities.

[0082] On the basis of the above, as a better implementation option, preferably, in this scheme S02, the working parameters of the crane include: driving current and braking torque; the parameter monitoring data includes driving current data and braking torque data; and the state monitoring data includes vibration data and noise data.

[0083] In order to improve the time consistency during data analysis, as a better implementation option, preferably, the data in the parameter monitoring data and the state monitoring data described in this solution are all attached with the timestamp when they were generated; their records are respectively expressed as follows:

[0084] S(t)=[S1(t),S2(t)]

[0085] Z9t)=[Z1(t),Z2(t)]

[0086] Where, t is time, S(t) is parameter monitoring data, S 1( t ) , S 2( t ) They are driving current data and braking torque data respectively; Z ( t ) is the condition monitoring data, Z 1( t ) , Z 2( t ) They are vibration data and noise data respectively.

[0087] In the judgment of data anomalies, as a better implementation option, preferably, in this solution S03, a judgment result is generated by comparing the status monitoring data with the corresponding threshold value. When one of the data in the status monitoring data exceeds the preset corresponding threshold value, a judgment result pointing to working abnormality is generated, and the corresponding characteristic working components with working abnormality are marked in the judgment result. Otherwise, a judgment result pointing to normal operation is generated.

[0088] In terms of multimodal fusion processing of data, as a preferred implementation option, preferably, in this solution S03, when the judgment result points to work abnormality, a multimodal fusion model based on environmental monitoring data, work monitoring data and status monitoring data is constructed, which is set as a multimodal feature matrix F(t), which is expressed as:

[0089] F(t)=[E(t),S(t),Z(t)]

[0090] Among them, E(t) is the environmental monitoring data, S(t) is the parameter monitoring data, and Z(t) is the state monitoring data;

[0091] The multimodal feature matrix F is trained by the random forest algorithm or neural network model. ( t) The parameter monitoring data, status monitoring data and environmental monitoring data are comprehensively analyzed to generate analysis results to distinguish whether the abnormal working state changes of the crane are induced by the internal factors of the crane and / or by the environmental factors; the internal factors are the working parameters of the crane, and the environmental factors are the environmental parameters when the crane is working.

[0092] As a preferred implementation option, preferably, in this solution S03, the parameter monitoring data, state monitoring data and environmental monitoring data in the multimodal feature matrix F(t) are comprehensively analyzed by a trained random forest algorithm, which includes:

[0093] The random forest algorithm is used to construct an anomaly classifier. The goal is to classify whether the abnormality is caused by the internal factors of the crane itself. The label Y is set and its value is assumed to be 0 or 1. 1 means the abnormality is caused by the internal factors of the crane itself, and 0 means no abnormality.

[0094] Construct K decision trees, and train each tree based on randomly sampled data from the training set (F(t), Y) until convergence;

[0095] Input the multimodal feature matrix F(t) to be detected, and let all decision trees give prediction results respectively

[0096] Aggregate the prediction results and use the voting method to output the final classification result, which is defined as follows:

[0097]

[0098] Among them, I(·) is the indicator function. When the assumption condition is met, that is, when the crane’s internal factors cause abnormal operation, its value is 1, otherwise it is 0; is the final prediction result, which is set to the analysis result, and k is the number of decision trees involved in the prediction.

[0099] The present application is not limited to the random forest algorithm in determining abnormal inducing factors. It can also use a trained neural network model to determine the multimodal feature matrix F. ( t ) The data in the crane are detected to obtain the judgment result that the change of the working state is induced by the internal factors of the crane and / or by the environmental factors. The training of the neural network is a common technology, and this embodiment will not be described in detail here.

[0100] As for the collection of abnormal data, in order to improve the convenience of subsequent reuse and the speed of retrieval and matching, as a better implementation option, preferably, in this solution S04, when the judgment result points to the abnormal operation of the crane, the environmental monitoring data, parameter monitoring data and status monitoring data of the preset time period before and after the abnormality occurs are collected and recorded, and the analysis results are then added with a timestamp and marked as abnormal information to form an abnormal data group, which is defined by the abnormal feature vector A in the following form:

[0101] A k =[E k , S k , Z k , t k ,y k ]

[0102] Among them, E k , S k , Z k , t k ,y k They are the environmental monitoring data, work monitoring data, state monitoring data, timestamp when the anomaly occurs, and anomaly information label at the time of the kth work anomaly respectively; finally, the anomaly feature vector A k Stored in the database.

[0103] Since there are many factors that may cause abnormalities when the crane is working, and this solution has a high reuse quality by collecting historical abnormal data, in order to avoid the risk of subsequent cranes being caught off guard under the same abnormal parameters, as a better implementation option, preferably, the monitoring method also includes:

[0104] S05. Acquire environmental monitoring data, work monitoring data and status monitoring data in real time, match them with abnormal data groups in the database, generate matching results, and generate risk warning information and output when the matching results meet the preset requirements.

[0105] Specifically, in solution S05 of this embodiment, the matching degree between the environmental monitoring data, the working monitoring data and the state monitoring data and the abnormal data group in the database is determined by similarity calculation, which includes:

[0106] The environmental monitoring data E current , work monitoring data S current Set the parameter F current =[E current , S current ], which corresponds to the abnormal feature vector A of the abnormal data group in the database k =[E k , S k The similarity of ] is defined as follows:

[0107]

[0108] Among them, α and β are weight coefficients, which are used to reflect the influence weight of the environment and the working state of the crane, and ||·|| is the Euclidean norm of the vector;

[0109] When the calculated similarity Sim(F current , A k ) exceeds the preset threshold λ, a risk warning message is generated and output.

[0110] After establishing a database containing abnormal data groups, this solution can match the real-time environmental monitoring data and work monitoring data with the abnormal data groups when monitoring the crane operation in the future, provide advance warnings when abnormalities may occur in the subsequent crane, and provide a reference for operators to avoid risks during operation.

[0111] Combination Figure 2 As shown, based on the above, this solution also proposes a crane operation control method, which is applied with the above-mentioned crane abnormality monitoring method, and further includes:

[0112] An abnormal data group is obtained, and according to the environmental monitoring data and parameter monitoring data therein, thresholds of the operating parameters of the crane under the corresponding environmental monitoring data and parameter monitoring data are adjusted.

[0113] In the crane work control method that applies the aforementioned crane abnormality monitoring method, the parameter thresholds of the crane can be adjusted under the working parameters where abnormalities have occurred to avoid the same abnormality from occurring again as much as possible, such as adjusting the torque output, adjusting the working current (indirectly reducing the drive output), etc. to improve the working reliability of the crane.

[0114] Combination Figure 3 As shown, based on the above, this solution also proposes a crane abnormality monitoring system, which includes:

[0115] The environmental monitoring module is deployed in the working area of ​​the crane and is used to collect environmental parameters and generate environmental monitoring data;

[0116] A monitoring sensor group is deployed on characteristic working parts of the crane and is used to collect vibration data and noise data when the crane is working to generate status monitoring data;

[0117] A working monitoring module is used to respond to a working start signal of the crane, record working parameters of the crane, and generate parameter monitoring data;

[0118] The data judgment module is used to obtain the status monitoring data in real time, judge the status monitoring data according to the preset conditions, and generate a judgment result to identify whether the crane is working abnormally;

[0119] The data fusion module is used to obtain the judgment result. When it points to the abnormal operation of the crane, a multimodal fusion model based on environmental monitoring data, work monitoring data and status monitoring data is constructed;

[0120] The data investigation module is used to comprehensively analyze the parameter monitoring data, the state monitoring data and the environmental monitoring data according to the preset conditions, and generate the analysis results to distinguish whether the abnormal working state of the crane is caused by the internal factors of the crane and / or by the environmental factors;

[0121] The data recording module is used to obtain the judgment result. When it points to abnormal operation of the crane, it collects and records the environmental monitoring data, parameter monitoring data and status monitoring data of the preset time period before and after the abnormality occurs, as well as the analysis results, and then attaches a timestamp and marks it as abnormal information to form an abnormal data group and store it in the database.

[0122] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0123] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, server, or network device, etc.) or a processor (processor) to perform all or part of the steps of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0124] The above descriptions are only some embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Any equivalent device or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for monitoring abnormalities of a crane, wherein the crane is deployed in an outdoor work site, an environmental monitoring module is deployed in the work site, and a monitoring sensor group is deployed on a characteristic working part of the crane, characterized in that: It includes: S01, the environmental monitoring module monitors the environmental parameters in the workplace in real time and generates environmental monitoring data; S02, responding to the work start signal of the crane, recording the working parameters of the crane and generating parameter monitoring data; at the same time, monitoring the state of the crane through the monitoring sensor group deployed on the crane and generating state monitoring data; S03, acquiring the status monitoring data in real time, judging the status monitoring data according to preset conditions, and generating a judgment result to identify whether the crane is working abnormally. When the crane is working abnormally, a multimodal fusion model based on the environmental monitoring data, the working monitoring data and the status monitoring data is constructed, and then the parameter monitoring data, the status monitoring data and the environmental monitoring data are comprehensively analyzed according to preset conditions to generate an analysis result to distinguish whether the working state change caused by the abnormal operation of the crane is induced by the internal factors of the crane and / or by the environmental factors; S04. Obtain the judgment result. When it points to abnormal operation of the crane, collect and record the environmental monitoring data, parameter monitoring data and status monitoring data of the preset time period before and after the abnormality occurs, as well as the analysis results, and then attach a timestamp and mark as abnormal information to form an abnormal data group, which is stored in the database.

2. The crane abnormality monitoring method according to claim 1, characterized in that: The environmental monitoring module includes one or more of a wind speed sensor, a humidity sensor, a temperature sensor, and a rain gauge sensor; In S01, the environmental monitoring data collected and generated by the environmental monitoring module is recorded in the form of a data set and marked with a relevant timestamp, and the record is represented as follows: E(t)=[E1(t)、E2(t)、…、E n (t)] Among them, t is the time point when the environmental monitoring module collects data, E1(t), E2(t), ..., E n (t) is the data item collected by the sensor of the environment monitoring module, which includes data collected by one or more sensors selected from the group consisting of a wind speed sensor, a humidity sensor, a temperature sensor, and a rain gauge sensor.

3. The crane abnormality monitoring method according to claim 2, characterized in that: The environmental monitoring data is also smoothed by the sliding average method to partially eliminate noise and abnormal data. The model is defined as follows: is the environmental data after sliding average processing, W is the sliding window size, which is used to represent the number of historical data points involved in the calculation; E i (j) is the collected value of the i-th environmental data at time j.

4. The crane abnormality monitoring method according to any one of claims 1 to 3, characterized in that: The crane is a bridge crane, and the characteristic working parts include a gantry, a luffing trolley, a hook drive motor, and a translation drive motor. The hook drive motor is deployed on the luffing trolley, and the translation drive motor is connected to both sides of the lower part of the gantry and is used to drive the gantry to translate on the track of the work site; The monitoring sensor group includes: a vibration sensor and / or a sound sensor, wherein the vibration sensor is used to collect vibration conditions of characteristic working parts of the crane when the crane is working, and the sound sensor is used to collect noise conditions of characteristic working parts of the crane when the crane is working; In S02, the operating parameters of the crane include: driving current and braking torque; The parameter monitoring data includes driving current data and braking torque data; The condition monitoring data includes vibration data and noise data; The data in the parameter monitoring data and the state monitoring data are all attached with the timestamp when they were generated; their records are respectively represented as follows: S(t ) =[S1(t)、S2(t)] Z(t)=[Z1(t),Z2(t)] Where t is time, S ( t ) is the parameter monitoring data, S1(t) and S2(t) are the driving current data and braking torque data respectively; Z(t) is the state monitoring data, Z1(t) and Z2(t ) They are vibration data and noise data respectively; In S03, a judgment result is generated by comparing the status monitoring data with the corresponding threshold value. When one of the data in the status monitoring data exceeds the preset corresponding threshold value, a judgment result pointing to working abnormality is generated, and the corresponding characteristic working parts where the working abnormality occurs are marked in the judgment result. Otherwise, a judgment result pointing to normal operation is generated.

5. The crane abnormality monitoring method according to claim 4, characterized in that: In S03, when the judgment result points to work abnormality, a multimodal fusion model based on environmental monitoring data, work monitoring data and status monitoring data is constructed, which is set as a multimodal feature matrix F(t), which is expressed as: F(t)=[E(t),S(t),Z(t)] Among them, E(t) is the environmental monitoring data, S(t) is the parameter monitoring data, and Z(t) is the state monitoring data; The parameter monitoring data, state monitoring data and environmental monitoring data in the multimodal feature matrix F(t) are comprehensively analyzed through the trained random forest algorithm or neural network model to generate analysis results to distinguish whether the working state changes caused by abnormal operation of the crane are induced by internal factors of the crane and / or by environmental factors; the internal factors are the working parameters of the crane, and the environmental factors are the environmental parameters when the crane is working.

6. The crane abnormality monitoring method according to claim 5, characterized in that: In S03, the parameter monitoring data, state monitoring data and environmental monitoring data in the multimodal feature matrix F(t) are comprehensively analyzed by the trained random forest algorithm, which includes: The random forest algorithm is used to construct an anomaly classifier. The goal is to classify whether the abnormality is caused by the internal factors of the crane itself. The label Y is set and its value is assumed to be 0 or 1. 1 means the abnormality is caused by the internal factors of the crane itself, and 0 means no abnormality. Construct K decision trees, and train each tree based on randomly sampled data from the training set (F(t), Y) until convergence; Input the multimodal feature matrix F(t) to be detected, and let all decision trees give prediction results respectively Aggregate the prediction results and use the voting method to output the final classification result, which is defined as follows: Among them, I(·) is the indicator function. When the assumption condition is met, that is, when the crane’s internal factors cause abnormal operation, its value is 1, otherwise it is 0; is the final prediction result, which is set to the analysis result, and k is the number of decision trees involved in the prediction.

7. The crane abnormality monitoring method according to claim 5, characterized in that: In S04, when the judgment result points to abnormal operation of the crane, the environmental monitoring data, parameter monitoring data and status monitoring data of the preset time period before and after the abnormality occurs are collected and recorded, and the analysis results are then added with a timestamp and marked as abnormal information to form an abnormal data group, which is defined by an abnormal feature vector A in the following form: A k =[E k ,S k ,Z k ,t k ,y k ] Among them, E k , S k , Z k , t k ,y k They are the environmental monitoring data, work monitoring data, state monitoring data, timestamp when the anomaly occurs, and anomaly information label at the time of the kth work anomaly respectively; finally, the anomaly feature vector A k Stored in the database.

8. The crane abnormality monitoring method according to claim 5, characterized in that: The monitoring method further comprises: S05. Real-time acquisition of environmental monitoring data, work monitoring data and status monitoring data, matching them with abnormal data groups in the database, generating matching results, and generating and outputting risk warning information when the matching results meet preset requirements; In S05, the matching degree between the environmental monitoring data, the working monitoring data and the state monitoring data and the abnormal data group in the database is determined by similarity calculation, which includes: The environmental monitoring data E current , work monitoring data S current Set the parameter F current =[E current , S current ], which corresponds to the abnormal feature vector A of the abnormal data group in the database k =[E k , S k The similarity of ] is defined as follows: Among them, α and β are weight coefficients, which are used to reflect the influence weight of the environment and the working state of the crane, and ||·|| is the Euclidean norm of the vector; When the calculated similarity Sim(F current , A k ) exceeds the preset threshold λ, a risk warning message is generated and output.

9. A crane operation control method, characterized in that: The method for monitoring abnormality of a crane according to any one of claims 1 to 8 is applied, and further comprises: Obtain an abnormal data group, and adjust the thresholds of the operating parameters of the crane under the corresponding environmental monitoring data and parameter monitoring data according to the environmental monitoring data and parameter monitoring data therein.

10. A crane abnormality monitoring system, characterized in that: It includes: The environmental monitoring module is deployed in the working area of ​​the crane and is used to collect environmental parameters and generate environmental monitoring data; A monitoring sensor group is deployed on characteristic working parts of the crane and is used to collect vibration data and noise data when the crane is working to generate status monitoring data; A working monitoring module is used to respond to a working start signal of the crane, record working parameters of the crane, and generate parameter monitoring data; The data judgment module is used to obtain the status monitoring data in real time, judge the status monitoring data according to the preset conditions, and generate a judgment result to identify whether the crane is working abnormally; The data fusion module is used to obtain the judgment result. When it points to the abnormal operation of the crane, a multimodal fusion model based on environmental monitoring data, work monitoring data and status monitoring data is constructed; The data investigation module is used to comprehensively analyze the parameter monitoring data, the state monitoring data and the environmental monitoring data according to the preset conditions, and generate the analysis results to distinguish whether the abnormal working state of the crane is caused by the internal factors of the crane and / or by the environmental factors; The data recording module is used to obtain the judgment result. When it points to abnormal operation of the crane, it collects and records the environmental monitoring data, parameter monitoring data and status monitoring data of the preset time period before and after the abnormality occurs, as well as the analysis results, and then attaches a timestamp and marks it as abnormal information to form an abnormal data group and store it in the database.

Images