An industrial park safety management method based on artificial intelligence

Abstract

The present invention discloses an industrial park safety management method based on artificial intelligence, which belongs to the field of industrial park management technology, and specifically comprises the following steps: counting the material types and corresponding standard material images of each production workshop, obtaining the route from the warehouse to the workshop, and recording the time, load, total height of materials and real-time material images of the transport vehicle leaving the warehouse and entering the workshop; calculating the estimated volume and estimated density of the material by using the total height of the material, and then calculating the estimated density range of the production material, and screening the production materials with density within the estimated density range, performing image comparison on the production materials that meet the density range to determine the type of production materials, and then selecting the production workshop of the corresponding production material type, so as to predict the driving route in advance and remind pedestrians to avoid it; the present invention realizes the monitoring of the transport vehicles traveling in the park, predicts the driving route in advance, and reminds pedestrians in the corresponding route to avoid it, thereby significantly reducing the risk of safety accidents.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial park management, and in particular to an industrial park safety management method based on artificial intelligence. Background Art

[0002] The application of artificial intelligence in industrial parks is extensive and in-depth, covering many aspects from safety management, energy management to traffic management and environmental monitoring. Through the deep integration of AI, the park's safety emergency management system has achieved a transformation from passive response to active prevention. The application of artificial intelligence in industrial parks is reflected in intelligent security monitoring, smart parking solutions, smart inspection systems, smart zero-carbon parks and energy management systems to promote energy conservation and emission reduction. These applications not only improve the operational efficiency and safety of the park, but also promote the park to develop in a more green, intelligent and efficient direction.

[0003] With the advent of the Industrial 4.0 era, industrial parks, as important places for production activities, have become particularly important for safety management. During the production activities in industrial parks, raw materials stored in warehouses are usually transported to production workshops by transport vehicles to meet the raw material needs of the production workshops. Transport vehicles need to travel frequently between the production workshops and warehouses.

[0004] However, since the roads in industrial parks are generally non-enclosed roads and there are many staff members in the park, there may be certain safety hazards to pedestrians during the driving of transport vehicles. Therefore, it is necessary to monitor the transport vehicles entering and leaving the warehouse in advance and determine the driving routes of the transport vehicles, so as to promptly remind pedestrians on the corresponding routes to avoid them. Summary of the invention

[0005] The purpose of this invention is to provide an industrial park safety management method based on artificial intelligence to solve the following technical problems:

[0006] Since the roads in industrial parks are generally non-enclosed roads and there are many staff members in the park, there may be certain safety hazards to pedestrians during the driving of transport vehicles. Therefore, it is necessary to monitor the transport vehicles entering and leaving the warehouse in advance and determine the driving routes of the transport vehicles, so as to promptly remind pedestrians on the corresponding routes to avoid them.

[0007] The purpose of the present invention can be achieved through the following technical solutions:

[0008] An industrial park safety management method based on artificial intelligence includes the following steps:

[0009] Statistics are collected on the types of production materials required by each production workshop in the industrial park, as well as the standard material images corresponding to each production material, and the driving routes from the production material storage warehouse to each production workshop are obtained; transportation information of the transport vehicle when it leaves the warehouse is collected, and the transportation information includes the time when the transport vehicle leaves the warehouse, the time when the transport vehicle enters the production workshop, the load, the total height of the material, and the real-time material image;

[0010] The estimated volume of the production materials loaded on the transport vehicle is calculated based on the length and width of the transport compartment and the total height of the materials. The estimated density ρ is calculated based on the load and the estimated volume. The density range is set to [ρ-a, ρ+a], where a is the preset threshold. The production materials that meet the density range are screened and marked as pending production materials. The standard material images corresponding to the pending production materials are compared with the real-time material images for similarity, and the type of production materials loaded on the transport vehicle is determined based on the similarity results.

[0011] According to the type of production materials loaded on the transport vehicle, the production workshop that meets the type of production materials is selected and marked as the target production workshop, and the driving route between the target production workshop and the warehouse is obtained; and it is marked as a warning route, and according to the time when the transport vehicle leaves the warehouse, pedestrians on the driving route are reminded to avoid it.

[0012] As a further solution of the present invention: the specific process of generating the standard material image corresponding to each production material is as follows:

[0013] The original images of the production materials are captured at multiple angles, the original images of the production materials are acquired, the original images of the production materials are denoised and contrast enhanced to obtain preprocessed images, characteristic elements of the preprocessed images are identified and annotated, the size of the annotated preprocessed images is adjusted, and the preprocessed images are calibrated as standard material images.

[0014] As a further solution of the present invention: the specific process of calculating the estimated volume of production materials loaded on the transport vehicle is:

[0015] The real-time total height of the transport vehicle is obtained through laser ranging, and the real-time total height of the transport vehicle is compared with the total height of the transport vehicle in an empty state to obtain the total height of the materials. The estimated volume of the production materials loaded on the transport vehicle is calculated based on the length of the transport compartment, the width of the transport compartment and the total height of the materials.

[0016] As a further solution of the present invention: when it is identified that the transport vehicle returns from the production workshop to the warehouse, the driving route between the production workshop and the warehouse is obtained and marked as a warning route. According to the time when the transport vehicle leaves the production workshop, pedestrians in the warning route are reminded to avoid it.

[0017] As a further solution of the present invention: when it is identified that the transport vehicle returns from the park gate to the warehouse, the driving route between the park gate and the warehouse is obtained and marked as a warning route. According to the time when the transport vehicle enters the park gate, pedestrians in the warning route are reminded to avoid it.

[0018] As a further solution of the present invention: when it is identified that the transport vehicle starts from the starting position of the warning route, pedestrians in the warning route are reminded to avoid it until the transport vehicle reaches the end position of the warning route.

[0019] As a further solution of the present invention: the warning modes include sound warning, light flashing and text message reminder.

[0020] As a further solution of the present invention: the specific process of similarity comparison is:

[0021] According to the real-time material image and the standard material image, a DINOv2 model is constructed, the real-time material image and the standard material image are processed, the processed real-time material image and the processed standard material image are input into the DINOv2 model, a feature vector of the real-time material image and a feature vector of the standard material image are obtained, the feature vector of the real-time material image and the feature vector of the standard material image are compared, and a cosine similarity value is calculated, the cosine similarity value ∈ [-1, 1], the similarity values ​​are compared, the largest similarity value is selected, and the standard material image corresponding to the similarity value is determined, then the type of the production material is the type of production material corresponding to the standard material image.

[0022] As a further solution of the present invention: when it is identified that the transport vehicle is departing from the warehouse, the personnel of the target production workshop are reminded to prepare to receive the production materials.

[0023] Beneficial effects of the present invention:

[0024] The present invention obtains the approximate loading height of the production materials loaded on the transport vehicle through laser ranging, and calculates the estimated volume of the loaded production materials based on this, and then infers the estimated density range of the production materials, so as to screen possible production materials according to the estimated density range, and directly perform image comparison on the production materials that meet the density range, effectively reducing the number of images and equipment computing power required for material similarity comparison; further, only by identifying the type of production materials loaded on the transport vehicle, and then judging the destination workshop of the transport vehicle, the next driving route of the transport vehicle is predicted in advance, and pedestrians in the driving route are reminded to avoid in time to ensure pedestrian safety, and the driving route is monitored in real time during transportation to ensure transportation efficiency. The present invention realizes intelligent safety management of industrial parks. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The present invention will be further described below in conjunction with the accompanying drawings.

[0026] Figure 1 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0027] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0028] See also Figure 1 As shown, the present invention is an industrial park safety management method based on artificial intelligence, comprising the following steps:

[0029] Statistics are collected on the types of production materials required by each production workshop in the industrial park, as well as the standard material images corresponding to each production material, and the driving routes from the production material storage warehouse to each production workshop are obtained; transportation information of the transport vehicle when it leaves the warehouse is collected, and the transportation information includes the time when the transport vehicle leaves the warehouse, the time when the transport vehicle enters the production workshop, the load, the total height of the material, and the real-time material image;

[0030] When the production materials are all plastic, the length of the conventional transport compartment is 2.1m, the width of the transport compartment is 1.7m, the total height of the transport vehicle in the empty state is 1.6m, the total height of the materials is 2.5m, and the load of the transport vehicle is 2950kg. According to the length and width of the transport compartment and the total height of the materials, the estimated volume of the production materials loaded on the transport vehicle is 3.2m 3 , thus the estimated density of the production material is calculated to be 0.92 g / cm 3 , set density range [0.7g / cm 3 , 1.2g / cm 3 ], screen the production materials that meet the density range and mark them as pending production materials, compare the standard material images corresponding to the pending production materials with the real-time material images for similarity, and determine the type of production materials loaded on the transport vehicle based on the similarity results;

[0031] According to the type of production materials loaded on the transport vehicle, the production workshop that meets the type of production materials is selected and marked as the target production workshop, and the driving route between the target production workshop and the warehouse is obtained; and it is marked as a warning route, and according to the time when the transport vehicle leaves the warehouse, pedestrians on the driving route are reminded to avoid it.

[0032] Since the roads in the industrial park are generally non-enclosed roads and there are many staff in the park, there may be certain safety hazards to pedestrians during the driving of transport vehicles. Therefore, it is necessary to monitor the transport vehicles entering and leaving the warehouse in advance and determine the driving routes of the transport vehicles, so as to promptly remind pedestrians on the corresponding routes to avoid them;

[0033] The present invention obtains the approximate loading height of the production materials loaded on the transport vehicle through laser ranging, and calculates the estimated volume of the loaded production materials based on this, and then infers the estimated density range of the production materials, so as to screen possible production materials according to the estimated density range, and directly perform image comparison on the production materials that meet the density range, effectively reducing the number of images and equipment computing power required for material similarity comparison; further, only by identifying the type of production materials loaded on the transport vehicle, and then selecting the production workshop corresponding to the production material, the next driving route of the transport vehicle is predicted in advance, and pedestrians in the driving route are promptly reminded to avoid it to ensure pedestrian safety, and the driving route is monitored in real time during transportation to ensure transportation efficiency.

[0034] In a preferred embodiment of the present invention, the specific process of generating the standard material image corresponding to each production material is as follows:

[0035] The original images of the production materials are captured at multiple angles, the original images of the production materials are acquired, the original images of the production materials are denoised and contrast enhanced to obtain preprocessed images, characteristic elements of the preprocessed images are identified and annotated, the size of the annotated preprocessed images is adjusted, and the preprocessed images are calibrated as standard material images.

[0036] In view of the complexity of production materials, the present invention adopts original images of production materials taken from multiple angles to ensure that the characteristics of materials are captured from different perspectives, improves image quality through denoising and contrast enhancement technology, and highlights image details, making key features in the image easier to identify and mark, and adjusts the size of the marked pre-processed image to generate high-quality standard material images, so that the comparison results are more accurate when comparing similarities.

[0037] In another preferred embodiment of the present invention, the specific process of calculating the estimated volume of the production material loaded on the transport vehicle is as follows:

[0038] The real-time total height of the transport vehicle is obtained by scanning with laser ranging technology. The real-time total height of the transport vehicle is compared with the total height of the transport vehicle in an empty state to obtain the total height of the materials. The estimated volume of the production materials loaded on the transport vehicle is calculated based on the length of the transport compartment, the width of the transport compartment and the total height of the materials.

[0039] The present invention adopts laser ranging technology to measure the height of a certain section of the transport vehicle, and quickly obtains the top height of the transport vehicle in a loaded state through a single laser point, and compares it with the height of the transport vehicle in an empty state, and calculates the height increment of the transport vehicle after loading the material; considering the uneven placement of production materials, accurate measurement of the volume of production materials requires high equipment and time costs, which does not meet the needs of actual production activities; therefore, the present invention only uses the top height of the production material collected by a single laser rangefinder to calculate the estimated height of the production material, so as to obtain the estimated volume and estimated density of the production material loaded on the transport vehicle, so as to reduce the number of production material types that need to be considered when comparing material similarities, and improve the efficiency of data processing.

[0040] It is worth noting that when a transport vehicle is identified as returning from the production workshop to the warehouse, the driving route between the production workshop and the warehouse is obtained and marked as a warning route. According to the time when the transport vehicle leaves the production workshop, pedestrians in the warning route are reminded to avoid it.

[0041] When a transport vehicle is identified as returning from the park gate to the warehouse, the driving route between the park gate and the warehouse is obtained and marked as a warning route. According to the time when the transport vehicle enters the park gate, pedestrians in the warning route are reminded to avoid it.

[0042] Since the roads in industrial parks are generally non-enclosed roads and there are many staff members in the parks, and considering that people and vehicles may shuttle in the industrial parks at any time, in order to reduce the occurrence of safety accidents, the present invention deploys an automatic monitoring system. When the transport vehicle departs from the production warehouse or the park gate, the moving direction and speed of the transport vehicle are automatically tracked to determine the route of the transport vehicle. The system marks it as a warning route and reminds pedestrians in the warning route to avoid it to ensure the safety of pedestrians.

[0043] In another preferred situation of this embodiment, when it is identified that the transport vehicle starts from the starting position of the warning route, pedestrians in the warning route are reminded to give way until the transport vehicle reaches the end position of the warning route.

[0044] Taking into account the frequent flow of people in the industrial park, the present invention introduces an automated monitoring system to monitor in real time the entire process of the transport vehicle starting from the starting point of the warning route until it arrives at the end point, and continuously sends out warning signals to remind pedestrians in the warning route to avoid it.

[0045] It is worth noting that the warning methods include sound warnings, flashing lights and text message reminders.

[0046] In another preferred embodiment of the present invention, the specific process of similarity comparison is:

[0047] According to the real-time material image and the standard material image, a DINOv2 model is constructed, the real-time material image and the standard material image are processed, the processed real-time material image and the processed standard material image are input into the DINOv2 model, a feature vector of the real-time material image and a feature vector of the standard material image are obtained, the feature vector of the real-time material image and the feature vector of the standard material image are compared, and a cosine similarity value is calculated, the cosine similarity value ∈ [-1, 1], the similarity values ​​are compared, the largest similarity value is selected, and the standard material image corresponding to the similarity value is determined, then the standard material image represents the corresponding production material type.

[0048] Due to different lighting and shooting angles, the real-time material image may not be clear. The present invention pre-processes the real-time material image and the standard material image through AutoImageProcessor, converts them into an input format acceptable to the model, and extracts features. The processed image is input into the DINOv2 model. The last_hidden_state of the model is usually used to obtain the feature representation of the image, and then the global feature vector of the entire image is obtained by averaging the features. The cosine similarity is used to calculate the similarity of the two feature vectors. The torch.nn.CosineSimilarity is usually used to implement this calculation. Since the value of cosine similarity is between -1 and 1, the closer the value is to 1, the higher the similarity. Therefore, the present invention compares the cosine similarity values, selects the largest similarity value, and determines the standard material image corresponding to the similarity value. The type of production material is the type of production material corresponding to the standard material image representation.

[0049] In another preferred embodiment of the present invention, when it is recognized that the transport vehicle has departed from the warehouse, personnel in the target production workshop are reminded to prepare to receive the production materials.

[0050] Since problems such as delayed transport vehicle scheduling and insufficient workshop preparation are often encountered during the production process, resulting in low production efficiency, in order to improve production efficiency, the present invention uses an automated monitoring system. When the transport vehicle departs from the production material warehouse, it automatically triggers the generation of an early warning route through real-time monitoring and data analysis, and notifies the personnel in the target production workshop in advance to prepare to receive the production materials. At the same time, it issues avoidance reminders to pedestrians along the route to improve production efficiency.

[0051] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. An industrial park safety management method based on artificial intelligence, characterized in that: The following steps are involved: Statistics are collected on the types of production materials required by each production workshop in the industrial park, as well as the standard material images corresponding to each production material, and the driving routes from the production material storage warehouse to each production workshop are obtained; transportation information of the transport vehicle when it leaves the warehouse is collected, and the transportation information includes the time when the transport vehicle leaves the warehouse, the time when the transport vehicle enters the production workshop, the load, the total height of the material, and the real-time material image; The estimated volume of the production materials loaded on the transport vehicle is calculated based on the length and width of the transport compartment and the total height of the materials. The estimated density ρ is calculated based on the load and the estimated volume. The density range is set to [ρ-a, ρ+a], where a is the preset threshold. The production materials that meet the density range are screened and marked as pending production materials. The standard material images corresponding to the pending production materials are compared with the real-time material images for similarity, and the type of production materials loaded on the transport vehicle is determined based on the similarity results. According to the type of production materials loaded on the transport vehicle, a production workshop that meets the type of production materials is selected and marked as the target production workshop, and the driving route between the target production workshop and the warehouse is obtained; It is marked as a warning route, and pedestrians on the route are reminded to give way based on the time when the transport vehicle leaves the warehouse.

2. According to the artificial intelligence-based industrial park safety management method of claim 1, it is characterized in that: The specific process of generating the standard material image corresponding to each production material is as follows: The original images of the production materials are captured at multiple angles, the original images of the production materials are acquired, the original images of the production materials are denoised and contrast enhanced to obtain preprocessed images, characteristic elements of the preprocessed images are identified and annotated, the size of the annotated preprocessed images is adjusted, and the preprocessed images are calibrated as standard material images.

3. According to claim 1, an industrial park safety management method based on artificial intelligence is characterized in that: The specific process for calculating the estimated volume of production materials loaded on a transport vehicle is: The real-time total height of the transport vehicle is obtained through laser ranging, and the real-time total height of the transport vehicle is compared with the total height of the transport vehicle in an empty state to obtain the total height of the materials. The estimated volume of the production materials loaded on the transport vehicle is calculated based on the length of the transport compartment, the width of the transport compartment and the total height of the materials.

4. According to the artificial intelligence-based industrial park safety management method of claim 1, it is characterized in that: When a transport vehicle is identified as returning from the production workshop to the warehouse, the driving route between the production workshop and the warehouse is obtained and marked as a warning route. According to the time when the transport vehicle leaves the production workshop, pedestrians in the warning route are reminded to avoid it.

5. The method for industrial park safety management based on artificial intelligence according to claim 1 is characterized in that: When a transport vehicle is identified as returning from the park gate to the warehouse, the driving route between the park gate and the warehouse is obtained and marked as a warning route. According to the time when the transport vehicle enters the park gate, pedestrians in the warning route are reminded to avoid it.

6. An artificial intelligence-based industrial park safety management method according to any one of claims 1, 4, and 5, characterized in that: When it is identified that the transport vehicle departs from the starting position of the warning route, pedestrians in the warning route are reminded to give way until the transport vehicle reaches the end position of the warning route.

7. The method for industrial park safety management based on artificial intelligence according to claim 6 is characterized in that: Warning methods include sound warnings, flashing lights and text message reminders.

8. The method for industrial park safety management based on artificial intelligence according to claim 1 is characterized in that: The specific process of similarity comparison is: According to the real-time material image and the standard material image, a DINOv2 model is constructed, the real-time material image and the standard material image are processed, the processed real-time material image and the processed standard material image are input into the DINOv2 model, a feature vector of the real-time material image and a feature vector of the standard material image are obtained, the feature vector of the real-time material image and the feature vector of the standard material image are compared, and a cosine similarity value is calculated, the cosine similarity value ∈ [-1, 1], the similarity values ​​are compared, the largest similarity value is selected, and the standard material image corresponding to the similarity value is determined, then the type of the production material is the type of production material corresponding to the standard material image.

9. The method for industrial park safety management based on artificial intelligence according to claim 1 is characterized in that: When the transport vehicle is identified as departing from the warehouse, the personnel in the target production workshop are reminded to prepare to receive the production materials.

Images