Cold source disaster-causing object intelligent identification and early warning method and device integrated with AI algorithm

Abstract

The invention relates to the technical field of cold source disaster-causing object detection, and discloses a cold source disaster-causing object intelligent identification and early warning method and device integrated with an AI algorithm. According to the cold source disaster-causing object intelligent identification and early warning method and device integrated with the AI algorithm, the characteristic parameters of the cold source disaster-causing object are dynamically regulated and controlled based on different temperature intervals, so that the false identification condition is effectively reduced, the accuracy of cold source disaster-causing object identification is improved, and the method and the device have the advantages that the identification efficiency is improved; collected data is converted into an input format suitable for the AI algorithm according to a preset format and a preprocessing mode, it is ensured that the AI algorithm can efficiently process the data, a foundation is laid for subsequent accurate analysis and prediction, a deep learning model is trained through the preprocessed data, a prediction result is output based on a set confidence coefficient threshold value, and the prediction accuracy is improved. And finally, the alarm threshold value can be dynamically adjusted according to the confidence coefficient in the prediction result, so that excessive sensitive alarm is avoided, the rationality and effectiveness of early warning are improved, and unnecessary alarm interference is reduced.

Description

Technical Field

[0001] This application relates to the technical field of cold source disaster-causing object detection, and particularly to an intelligent recognition and early warning method and device for cold source disaster-causing objects integrating AI algorithms. Background Art

[0002] The intelligent recognition and early warning method and device for cold source disaster-causing objects integrating AI algorithms is a technical means that realizes real-time monitoring and automatic early warning by combining artificial intelligence technology, and is specifically used to identify potential dangerous objects (such as pipeline leaks, cold source equipment failures, etc.) in low-temperature environments. This device uses deep learning algorithms to process data streams from multiple sensors and quickly analyzes and identifies possible cold source disaster-causing situations, providing a scientific basis for timely taking preventive measures. However, this technology currently also faces several problems that need to be solved. First, how to accurately regulate the characteristic parameters of cold source disaster-causing objects in different temperature ranges is one of the key challenges. Since the change of environmental temperature may lead to misidentification phenomena, affecting the accuracy of judgment. Second, optimizing the data input format and preprocessing method for AI algorithms is crucial. Data noise will significantly affect the model prediction accuracy, and an effective noise reduction mechanism needs to be found. Finally, adjusting the confidence threshold of the deep learning model is also very necessary. Setting it too low may cause too many false positive alarms, while setting it too high may miss important warning signals. Therefore, finding a suitable confidence threshold to balance the problem of overly sensitive alarms is an important consideration. Summary of the Invention

[0003] The embodiments of the present disclosure provide an intelligent recognition and early warning method and device for cold source disaster-causing objects integrating AI algorithms, which at least partially solve the problems existing in the prior art.

[0004] The intelligent recognition and early warning method for cold source disaster-causing objects integrating AI algorithms includes:

[0005] S101. Dynamically regulate the characteristic parameters of cold source disaster-causing objects based on different temperature ranges to reduce misidentification situations;

[0006] S102. Collect the data after the regulation of characteristic parameters and convert it into an input format suitable for the AI algorithm according to a preset format and preprocessing method;

[0007] S103. Train a deep learning model with the preprocessed data and output a prediction result based on a set confidence threshold;

[0008] S104. Dynamically adjust the alarm threshold according to the confidence in the prediction result to prevent overly sensitive alarms.

[0009] Dynamically regulating the characteristic parameters of cold source disaster-causing objects based on different temperature ranges specifically includes the following steps:

[0010] Divide the ambient temperature into several different intervals, and define corresponding characteristic parameter adjustment strategies for each interval;

[0011] Collect the actual temperature distribution of each temperature zone, and determine the temperature threshold Tk (where k represents the kth temperature zone);

[0012] In each temperature zone, determine the new characteristic parameter P = Pbase + ΔP according to the following formula, where ΔP is the temperature zone specific adjustment value;

[0013] If the ambient temperature T ∈ [Tk-1, Tk) and the error E < Tmax, then apply the above characteristic parameter P to provide data to the model; where Pbase is the base parameter, Tk-1 and Tk represent the temperature zone range boundaries, and E represents the difference between the actual measurement and the preset parameter.

[0014] When dynamically updating the characteristic parameters of the cold source disaster-causing substances based on the specific conditions of different temperature zones, the following steps are further included:

[0015] Determine the initial parameter reference value;

[0016] Calculate the variance S of the data collected in each temperature zone;

[0017] Based on the change trend of the variance S in each interval, adjust the parameter adjustment amplitude A(S) = α * sqrt(S) of the corresponding interval;

[0018] If the current variance increases significantly (for example, S > δ), then increase A, otherwise keep it unchanged, where α represents the gain coefficient and δ represents the determination threshold.

[0019] When dynamically calibrating the characteristic parameters based on the data collected in each degree of temperature zone, the following operations should also be included:

[0020] Regularly record and summarize the temperature data over a period of time;

[0021] Calculate the maximum and minimum temperature ranges D = Tmax(k) - Tmin(k) in each interval during this period;

[0022] Analyze the maximum error Emax caused by internal and external factors in each interval;

[0023] If the range exceeds a certain limit and Emax exceeds the set threshold, then based on C = max(D / Emax, γ), strengthen the key monitoring efforts for these abnormal areas, that is, increase the sensitivity of the characteristic parameters in this section. C refers to the control factor and γ refers to the preset minimum ratio.

[0024] After adjusting the monitoring focus based on the temperature range evaluation results, it is further refined into the following process:

[0025] Implement a periodic review and calibration mechanism for each temperature zone;

[0026] Comprehensively consider the data deviation accumulated during long-term operation;

[0027] Design targeted solutions for places where specific problems frequently occur;

[0028] When it is found that the misrecognition rate in a certain interval repeatedly exceeds r or the error of three consecutive samples is higher than b, start the emergency plan processing mode for this temperature zone, where r is the tolerance error ratio and b represents the upper limit of the allowable deviation range.

[0029] The process of optimizing the alarm sensitivity based on the misrecognition frequency statistics data specifically includes:

[0030] Statistically calculate the false alarm probability Fp and the missed alarm ratio Fneg in each temperature range;

[0031] Check all potential dangerous situations and their possible impact severity levels;

[0032] Predict the possibility of potential disaster events;

[0033] Comprehensively evaluate the early warning quality of different intervals through Q=(Fp / w1)+(Fneg / w2), and adopt more stringent standards for areas with higher Q. w1 and w2 are weight factors used to adjust the different proportions of their contributions.

[0034] In terms of filtering the influence of data noise, in addition to the previous methods, the following improvement points are also included:

[0035] Apply adaptive filtering technology to process the original sensor output data;

[0036] Design a special denoising function to remove random noise components;

[0037] Establish a more reliable preprocessing process using historical experience knowledge;

[0038] Assume that the signal-to-noise ratio SNR < smin (smin represents the lowest required signal-to-noise ratio setting), and the acquisition scheme should be readjusted to improve the data quality. SNR is a quantitative indicator of the relative intensity comparison between the signal and the noise.

[0039] More detailed specifications are made on how to effectively reduce the unnecessary complexity in the model input:

[0040] Establish a feature selection framework to select the subset of dimensions that can best represent the target characteristics;

[0041] Reduce the proportion of useless information and improve the training efficiency;

[0042] Reduce the dimension using sparse coding while keeping the core attributes intact;

[0043] When the remaining feature dimension d is greater than the given threshold Nt and the cross-validation accuracy ACC decreases, appropriately reduce the value of d until the optimal performance level is met.

[0044] To prevent the instability and uncertainty caused by the model relying too much on certain key but unstable factors during the prediction process, the following strategies are proposed:

[0045] Build a stability assessment module to monitor the fluctuations during each iteration;

[0046] Conduct a comprehensive scan of all feature vectors to find weak links;

[0047] Strengthen the association strength between important nodes to ensure the robustness of the network;

[0048] If the change amplitude of any parameter Xn at any moment is greater than ∈ or the change directions are the same for multiple consecutive frames, trigger the safety inspection program to confirm whether to continue learning and updating along this path, where ∈ represents the change amplitude limit standard, which reflects the tolerance for system stability.

[0049] On the basis of continuously optimizing the confidence score, the adaptability response of the system is further enhanced, and this system is specifically embodied as:

[0050] Implement multiple rounds of experimental tests to obtain the optimal performance parameter configuration under various environments;

[0051] Continuously collect actual usage feedback to guide the next iteration optimization direction;

[0052] Customize personalized service options according to user needs to enhance the user experience;

[0053] When the confidence Conf after each prediction is lower than the critical threshold θ, automatically reduce the urgency level L of the next alarm until it returns to the normal state after multiple high-accuracy verifications are accumulated.

[0054] A cold source disaster-causing object intelligent identification and early warning device integrating AI algorithms, the device performs operations through the cold source disaster-causing object intelligent identification and early warning method integrating AI algorithms, and the system includes:

[0055] Regulation module: Responsible for managing and coordinating various data acquisition devices, and dynamically adjusting the working parameters of the acquisition devices according to different environmental conditions, monitoring requirements, etc., such as the shooting frequency of underwater cameras, the detection range and frequency of sonar devices, etc., to ensure that high-quality, comprehensive and cold source disaster-causing object identification and early warning-related data are collected.

[0056] Conversion module: Convert some non-numerical data, such as text descriptions, category labels, etc., into numerical data suitable for AI algorithms. For example, convert the type name of cold source disaster-causing substances into corresponding digital codes, or convert the color information in images into numerical vectors, so that the AI model can calculate and learn these data.

[0057] Learning module: Automatically extract key features from a large amount of cold source disaster-causing substance-related data collected by the data acquisition module. For example, for the image data captured by an underwater camera, the learning module can learn visual features such as the shape, texture, and color of cold source disaster-causing substances through technologies such as convolutional neural networks; for sonar data, it can learn acoustic features such as the reflection intensity and echo time of the target object, thereby converting these complex raw data into a feature table that is more valuable for recognition and early warning.

[0058] In summary, the beneficial technical effects of this application are as follows:

[0059] 1. The intelligent recognition and early warning method and device for cold source disaster-causing substances integrating AI algorithms effectively reduce the misrecognition situation and improve the accuracy of cold source disaster-causing substance recognition by dynamically adjusting the characteristic parameters of cold source disaster-causing substances based on different temperature ranges.

[0060] 2. The intelligent recognition and early warning method and device for cold source disaster-causing substances integrating AI algorithms convert the collected data into an input format suitable for AI algorithms according to the preset format and preprocessing method, ensuring that the AI algorithm can efficiently process the data and laying a foundation for subsequent accurate analysis and prediction.

[0061] 3. The intelligent recognition and early warning method and device for cold source disaster-causing substances integrating AI algorithms use the preprocessed data to train a deep learning model and output prediction results based on the set confidence threshold, making the prediction have a certain degree of reliability and scientificity.

[0062] 4. The intelligent recognition and early warning method and device for cold source disaster-causing substances integrating AI algorithms dynamically adjust the alarm threshold according to the confidence level in the prediction results, avoid overly sensitive alarms, improve the rationality and effectiveness of early warning, and reduce unnecessary alarm interference. Brief Description of the Drawings

[0063] Figure 1 is the flowchart of the intelligent recognition and early warning method for cold source disaster-causing substances integrating AI algorithms of the present invention.

[0064] Figure 2 is the flowchart of the intelligent recognition and early warning system for cold source disaster-causing substances integrating AI algorithms of the present invention. Detailed Embodiments

[0065] The embodiments of the present application will be described in detail below, and examples of the embodiments are shown in the drawings.

[0066] In the description of this specification, the descriptions referring to the terms "certain embodiments", "one embodiment", "some embodiments", "schematic embodiments", "examples", "specific examples", or "some examples" mean that the specific features, structures, materials, or characteristics described in connection with the embodiments or examples are included in at least one embodiment or example of the present application. In this specification, the schematic expressions of the above terms do not necessarily refer to the same embodiments or examples. Moreover, the specific features, structures, materials, or characteristics described can be combined in a suitable manner in any one or more embodiments or examples.

[0067] Referring to the attached Figure 1 drawings, a method for intelligent identification and early warning of cold source disaster-causing substances integrated with AI algorithms of the present invention is described. The method includes dynamically regulating the characteristic parameters of cold source disaster-causing substances within different temperature ranges to reduce the possibility of misidentification. By adjusting the characteristic parameters, the data changes under different environmental conditions can be effectively adapted. At the same time, after the data is collected and undergoes strict formatting and preprocessing steps, it is converted into a format suitable for input into the deep learning model. On this basis, the trained model outputs a prediction result based on a preset confidence threshold. Finally, according to the confidence in the obtained prediction result, the dynamic adjustment of the alarm threshold is implemented to avoid over-sensitive alarms.

[0068] Based on different temperature ranges, the system first dynamically regulates the characteristic parameters of cold source disaster-causing substances. This means that in different working environments (such as high-temperature areas, normal temperature, and low-temperature areas), the environmental status is monitored through devices such as sensors, and then the key parameters that are prone to misunderstanding or misidentification in specific situations are optimized, such as the shape of the object, the thermal characteristics of the material, and the surface temperature fluctuation. These adjustments ensure that high-accuracy data can still be obtained in a complex and changeable environment and input into the next-level processing module.

[0069] During the specific operation process, adjustment rules on multiple sub-dimensions will be set according to different categories of targets, and an intelligent judgment mechanism will be formed by combining historical experience data accumulation to update the weights of each factor in real time. For example, in a chemical production workshop, there are various chemical medicine tanks stored in rooms with different insulation measures, and the environments they are in may vary greatly from indoor and outdoor extreme weather. For this scenario, adjustment formulas for the associated attributes of each chamber type can be built into the software system. During actual detection, it can quickly respond to environmental changes and automatically correct relevant physical indicators to make the subsequent processing more accurate and efficient.

[0070] After completing the initial parameter adjustment, the device will reorganize and classify all data streams from on-site monitoring sites in a unified planned format and perform necessary signal cleaning measures to meet the information requirements of downstream algorithms, that is, convert the original collected data into an expression form that meets specific artificial intelligence technical specifications in a predefined format and method, such as removing redundant parts and retaining only key elements that help identify analysis, or using mathematical transformations to eliminate certain periodic disturbance components to ensure consistency and stability between data samples. This directly improves the final output quality, reduces the computational burden, and enhances the model's robustness and generalization capabilities.

[0071] The clean and normalized data set processed by the above method can be used to train advanced statistical classifiers such as deep neural networks with self-learning capabilities, so that they can gradually master the ability to distinguish normal situations from other potential hazards. During the entire iterative evolution period, new batches of labeled samples must be continuously added to the training set to expand coverage and improve feature extraction logic until the ideal performance level is reached before it can be put into formal use. At this stage, a large number of trial runs will be involved to evaluate the model's accuracy and recall performance to find the best configuration point. In particular, attention should be paid to preventing problems such as too much or too little deviation or complete mismatch.

[0072] Once the complete finished version that has been strictly debugged and verified is launched, it will use the credible score threshold determined in the early stage as a trigger condition to issue risk warning information to the user-side platform to notify relevant personnel to take emergency evacuation measures to curb the trend of worsening spread and loss. However, relying solely on fixed boundary judgments will cause frequent triggering of false alarms and trouble daily management affairs. Therefore, it is necessary to introduce flexible response strategies to establish a dynamic decision-making model through the study of the statistical laws of past accumulated cases. That is, as the system accumulates stable operating statistical data over a sufficiently long period of time, it supports a series of probability estimation functions to guide the single prediction score in the current scenario. Within which range, it is more reasonable and scientific to adjust the alarm frequency. This not only protects the system's keen touch but also flexibly balances the cost-benefit relationship, making the entire early warning mechanism more practical and closer to the needs of the application.

[0073] In one embodiment, it is assumed that signs of ice blockage are found inside the air-conditioning chilled water pipeline of a certain subway station, and this innovative technology needs to be launched to carry out a comprehensive diagnostic inspection activity. In the initial stage, the engineer team first needs to collect information on the microclimate status and the types and concentrations of substances carried around each node position in the chilled water supply chain within the station area, and then, with the help of professional and background knowledge, deduce the expected problem characteristic curve graph, and then formulate a set of detailed adjustment criteria by referring to theoretical literature. Immediately afterwards, professional measuring instruments are used to sequentially scan and record readings along the pipeline path and generate visual tables and charts for further comparison, reference and calibration. Subsequently, according to the pre-written data cleaning process, abnormal spikes and rough edges are removed to maintain the overall smoothness for the efficient operation of the model training and recognition process. After several rounds of careful polishing and optimization, a highly reliable and accurate report is finally obtained to help the operation department make various preventive measures in advance to ensure that the safe and smooth travel environment of citizens is not affected by interference and threats.

[0074] The intelligent identification and early warning method and device for cold source disaster-causing substances integrated with AI algorithms of the present invention include:

[0075] First of all, the overall steps involve data collection of cold source disaster-causing substances, dynamic regulation of characteristic parameters, data preprocessing, deep learning model training, confidence threshold setting and alarm control. Specifically, the system will comprehensively and real-time monitor cold source disaster-causing substances (such as leakage gas or ice blockage phenomenon) in different temperature ranges through a variety of sensors, and adjust the characteristic parameters according to the specific conditions of the environment to reduce the possibility of misidentification.

[0076] In order to solve the problem of how to regulate the characteristic parameters of cold source disaster-causing substances in different temperature ranges to solve the problem of misidentification, the system will automatically select the most suitable characteristic regulation method according to the different temperature environments during actual operation, that is, the system has a series of characteristic regulation parameter combination schemes corresponding to various temperature conditions, and these schemes can effectively reduce the misidentification rate caused by the uncertainty brought by temperature changes. For example, for problems such as the solidification of special materials generated in low-temperature environments, the system can enhance the effective identification of target substances under this condition and filter out non-disaster-causing temperature-related fluctuations by increasing the specific band spectrum or improving the detection sensitivity of certain physical and chemical properties.

[0077] Next, regarding how to regulate the data input format and preprocessing method of the AI algorithm to solve the problem that data noise affects the accuracy of the model, the present invention designs proprietary data conversion and preprocessing logic to ensure that the original monitoring data is converted into a standardized and denoised model input form. This process is divided into two steps: First, after collecting the characteristic parameters, perform normalization processing on them - this step ensures that all types of data are compared on the same scale, eliminating the interference caused by dimensional differences; Second, deeply process the possible data noise problems, introduce an adaptive filter and a feature enhancement algorithm to remove or weaken the influence of abnormal mutation values. In addition, according to the characteristics of the cold source scenario, a redundant verification link is specially set up to cross-compare the consistency and reliability of multiple measurement results, so as to further optimize the model training quality and ensure the robustness of the prediction results.

[0078] Finally, regarding how to regulate the confidence threshold of the deep learning model to solve the problem of over-sensitive alarms, the present invention proposes a method of dynamically setting warning limits based on a real-time feedback mechanism. If the traditional fixed confidence level is too low, it is easy to cause frequent false alarms and disturb the people, and if the quality standard is too high, it may miss the real crisis moment. Our solution combines three comprehensive indicators: actual environmental variable factors, past case learning experience, and current system performance evaluation, to achieve precise configuration of the best alarm sensitivity. In this way, under normal operating conditions, a relatively strict threshold can be set to avoid unnecessary warnings; while when a potential high risk is detected, quickly and flexibly lower the alarm threshold and start the emergency response procedure in a timely manner. In addition, a self-check and review mechanism is established to regularly review and analyze false alarm situations and fine-tune parameters, forming a benign self-evolution closed-loop system, so as to ensure that the device maintains high-efficiency and stable operation efficiency in the long term and effectively meet the user's safety protection needs.

[0079] In one embodiment, a cold source disaster-causing object intelligent identification and early warning device integrated with an AI algorithm is also disclosed. This evaluation system is executed by the above-mentioned cold source disaster-causing object intelligent identification and early warning method integrated with an AI algorithm, as Figure 2 shown, this system includes:

[0080] Regulation module: Responsible for managing and coordinating various data collection devices, and dynamically adjusting the working parameters of the collection devices according to different environmental conditions, monitoring requirements, etc., such as the shooting frequency of underwater cameras, the detection range and frequency of sonar devices, etc., to ensure that high-quality, comprehensive data related to the identification and early warning of cold source disaster-causing objects is collected.

[0081] Conversion Module: Convert some non-numerical data, such as text descriptions, category labels, etc., into numerical data suitable for AI algorithms. For example, convert the type name of cold source disaster-causing substances into corresponding digital codes, or convert the color information in images into numerical vectors, enabling the AI model to perform calculations and learning on this data.

[0082] Learning Module: Automatically extract key features from a large amount of cold source disaster-causing substance-related data collected by the data acquisition module. For example, for image data captured by an underwater camera, the learning module can learn visual features such as the shape, texture, and color of cold source disaster-causing substances through technologies such as convolutional neural networks; for sonar data, it can learn acoustic features such as the reflection intensity and echo time of the target object, thus converting these complex raw data into a feature table that is more valuable for recognition and early warning.

[0083] Each of the processes and treatments described above, such as a method, can be executed by a CPU. For example, in some embodiments, a method can be implemented as a computer software program that is tangibly contained in a machine-readable medium, such as a memory. In some embodiments, part or all of the computer program can be loaded and / or installed onto a computing device via a ROM and / or a communication unit. When the computer program is loaded into the RAM and executed by the CPU, it can perform a method for intelligent monitoring of cold source disaster-causing substances based on multi-sensor fusion or multiple actions described above.

[0084] The present disclosure relates to methods, computing devices, computer-readable storage media, and / or computer program products. The computer program product can include computer-readable program instructions for performing various aspects of the present disclosure.

[0085] A computer-readable storage medium can be a tangible device that can hold and store instructions for use by an instruction execution device. A computer-readable storage medium can be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the computer-readable storage medium include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disc read-only memory (CD-ROM), a digital versatile disc (DVD), a memory stick, a floppy disk, a mechanically encoded device such as a punched card or raised structures in grooves having instructions stored thereon, and any suitable combination of the foregoing. The computer-readable storage medium as used herein is not construed to be a transitory signal per se, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagating through a waveguide or other transmission medium (e.g., an optical pulse through an optical fiber cable), or an electrical signal transmitted through a wire.

[0086] The computer-readable program instructions described herein can be downloaded from a computer-readable storage medium to respective computing / processing devices, or can be downloaded to an external computer or an external storage device through a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network can include a copper transmission cable, an optical fiber transmission, a wireless transmission, a router, a firewall, a switch, a gateway computer, and / or an edge computing device. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in a computer-readable storage medium in each computing / processing device.

[0087] The computer program instructions for performing the operations of the present disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine - related instructions, microcode, firmware instructions, state - setting data, or source code or object code written in any combination of one or more programming languages, including object - oriented programming languages such as Smalltalk, C++, etc., and conventional procedural programming languages such as the "C" language or similar programming languages. The computer - readable program instructions may be executed entirely on the user's computer, partially on the user's computer, executed as a stand - alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider). In some embodiments, by using the state information of the computer - readable program instructions to customize an electronic circuit, such as a programmable logic circuit, a field - programmable gate array (FPGA), or a programmable logic array (PLA), the electronic circuit can execute the computer - readable program instructions to implement various aspects of the present disclosure.

[0088] These computer - readable program instructions can be provided to a processing unit of a general - purpose computer, a special - purpose computer, or other programmable data - processing apparatus to produce a machine, such that when these instructions are executed by the processing unit of the computer or other programmable data - processing apparatus, a device is produced that implements the functions / actions specified in one or more blocks of the flowchart and / or block diagram. These computer - readable program instructions can also be stored in a computer - readable storage medium, and these instructions cause a computer, a programmable data - processing apparatus, and / or other devices to work in a specific manner, so that the computer - readable medium storing the instructions includes a manufactured article that includes instructions for implementing various aspects of a multi - sensor - fusion - based intelligent cold - source disaster - causing object monitoring method or multiple actions.

[0089] The computer - readable program instructions can also be loaded onto a computer, other programmable data - processing apparatus, or other devices, such that a series of operation steps are executed on the computer, other programmable data - processing apparatus, or other devices to produce a computer - implemented process, so that the instructions executed on the computer, other programmable data - processing apparatus, or other devices implement a multi - sensor - fusion - based intelligent cold - source disaster - causing object monitoring method or multiple actions.

[0090] The embodiments of the present disclosure have been described above. The above description is exemplary and not exhaustive, and is also not limited to the disclosed embodiments. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, the practical application, or the improvement of technologies in the market, or to enable other ordinary skilled persons in the art to understand the embodiments disclosed herein.

[0091] Based on the same inventive concept, an embodiment of the present invention provides an intelligent terminal, including a memory and a processor, and a computer program capable of being loaded and executed by the processor is stored on the memory.

[0092] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present application. Those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present application.

Claims

1. An intelligent identification and early warning method for cold source disaster-causing objects integrating AI algorithm, characterized in that: Including: S101. Dynamically regulate the characteristic parameters of cold source disaster-causing substances based on different temperature ranges to reduce misidentification; S102. Collect the data after the regulation of characteristic parameters, and convert it into the input format suitable for AI algorithms according to the preset format and preprocessing method; S103. Train a deep learning model with the preprocessed data, and output the prediction result based on the set confidence threshold; S104. Dynamically adjust the alarm threshold according to the confidence level in the prediction result to prevent over-sensitive alarms.

2. The method for intelligent identification and early warning of cold source disaster-causing objects integrated with AI algorithm according to claim 1 is characterized in that: Dynamically regulating the characteristic parameters of cold source disaster-causing substances based on different temperature ranges specifically includes the following steps: Divide the ambient temperature into several different ranges, and define the corresponding characteristic parameter adjustment strategies for each range; Collect the actual temperature distribution of each temperature zone, and determine the temperature threshold Tk; In each temperature zone, determine the new characteristic parameter P = Pbase + ΔP according to the following formula, where ΔP is the temperature zone specific adjustment value; If the ambient temperature T ∈ [Tk - 1, Tk) and the error E < Tmax, then apply the above characteristic parameter P to provide data to the model.

3. The method for intelligent identification and early warning of cold source disaster-causing objects integrated with AI algorithm according to claim 2 is characterized in that: When dynamically updating the characteristic parameters of cold source disaster-causing substances based on the specific conditions of different temperature zones, it further includes the following steps: Determine the initial parameter reference value; Calculate the variance S of the data collected in each temperature zone; Based on the change trend of the variance S in each interval, adjust the parameter adjustment amplitude A(S) = α * sqrt(S) of the corresponding interval; If the current variance increases significantly, then increase A, otherwise keep it unchanged, where α represents the gain coefficient and δ represents the decision threshold.

4. The method for intelligent identification and early warning of cold source disaster-causing objects integrated with AI algorithm according to claim 3 is characterized in that: When dynamically calibrating the characteristic parameters based on the data collected in each degree of temperature zone, the following operations should also be included: Regularly record and summarize the temperature data for a period of time; Calculate the maximum and minimum temperature ranges D = Tmax(k) - Tmin(k) in each interval during this period; Analyze the maximum error Emax caused by internal and external factors in each interval; If the range exceeds a certain limit and Emax exceeds the set threshold, then according to C = max(D / Emax, γ), strengthen the key monitoring intensity of these abnormal areas, that is, increase the sensitivity of the characteristic parameters in this section.

5. The method for intelligent identification and early warning of cold source disaster-causing objects integrated with AI algorithm according to claim 4 is characterized in that: After adjusting the monitoring focus based on the temperature range evaluation results, it is further refined into the following process: Implement a periodic review and correction mechanism for each level of temperature zone; Comprehensively consider the data deviation accumulated during long-term operation; Design targeted solutions for places where specific problems frequently occur; When it is found that the misidentification rate in a certain interval repeatedly exceeds r or the error of three consecutive samples is higher than b, start the emergency plan processing mode for this temperature zone, where r is the tolerance error ratio and b represents the upper limit of the allowable deviation range.

6. The method for intelligent identification and early warning of cold source disaster-causing objects integrated with AI algorithm according to claim 5 is characterized in that: In the process of optimizing the alarm sensitivity based on the misidentification frequency statistics data, it particularly includes: Statistically calculate the false alarm probability Fp and the missed alarm ratio Fneg in each temperature range; Check all potential dangerous situations and their possible impact severity levels; Predict the possibility of potential disaster events; Comprehensively evaluate the early warning quality of different intervals through Q = (Fp / w1) + (Fneg / w2), and adopt more stringent standards for areas with higher Q. w1 and w2 are weighting factors used to adjust the different proportions of their contributions.

7. The method for intelligent identification and early warning of cold source disaster-causing objects integrated with AI algorithm according to claim 6 is characterized in that: In terms of filtering the influence of data noise, in addition to the previous methods, the following improvement points are included: Apply adaptive filtering technology to process the original data output by sensors; Design a special denoising function to remove random noise components; Utilize historical experience and knowledge to establish a more reliable preprocessing process; Assume that the signal-to-noise ratio SNR < smin, and the acquisition scheme should be readjusted to improve data quality. SNR is a quantitative indicator of the relative intensity comparison between the signal and the noise.

8. The method for intelligent identification and early warning of cold source disaster-causing objects integrated with AI algorithm according to claim 7 is characterized in that: More detailed specifications are made regarding how to effectively reduce the unnecessary complexity in the model input: Establish a feature selection framework to select the subset of dimensions that can best represent the target characteristics; Reduce the proportion of useless information to improve the training efficiency; Use sparse coding to reduce dimensions while keeping the core attributes intact; When the remaining feature dimension d is greater than the given threshold Nt and the cross-validation accuracy ACC decreases, appropriately reduce the value of d until the best performance level is met.

9. The method for intelligent identification and early warning of cold source disaster-causing objects integrated with AI algorithm according to claim 8 is characterized in that: To prevent the instability and uncertainty caused by the model relying too much on certain key but unstable factors during the prediction process, the following strategies are proposed: Construct a stability evaluation module to monitor the fluctuations during each iteration; Conduct a comprehensive scan of all feature vectors to find weak links; Strengthen the correlation strength between important nodes to ensure the robustness of the network; If the change amplitude of any parameter Xn at any moment is greater than ∈ or the change directions are the same for multiple consecutive frames, trigger the safety inspection program to confirm whether to continue learning and updating along this path. Here, ∈ represents the change amplitude limit standard, which reflects the tolerance for system stability; On the basis of continuously optimizing the confidence score, the adaptability response of the system is further enhanced. This system is specifically embodied as: Implement multiple rounds of experimental tests to obtain the optimal performance parameter configuration under various environments; Continuously collect actual usage feedback to guide the next iteration optimization direction; Customize personalized service options according to user needs to enhance the user experience; For each prediction where the confidence Conf is lower than the critical threshold θ, automatically reduce the urgency level L of the next alarm until it returns to the normal state after multiple high-accuracy verifications are accumulated.

10. An intelligent identification and early warning device for cold source disaster-causing objects integrating AI algorithm, characterized in that: The device performs operations through the intelligent identification and early warning method for cold source disaster-causing objects integrated with the AI algorithm described in any one of claims 1-9. The system includes: Regulation module: Responsible for managing and coordinating various data acquisition devices, and dynamically adjusting the working parameters of the acquisition devices according to different environmental conditions, monitoring requirements, etc., such as the shooting frequency of underwater cameras, the detection range and frequency of sonar devices, etc., to ensure the acquisition of high-quality, comprehensive data related to the identification and early warning of cold source disaster-causing objects; Conversion module: converts some non-numeric data, such as text descriptions, category labels, etc., into numerical data suitable for AI algorithm processing. For example, converts the type name of cold source disaster-causing objects into corresponding digital codes, or converts the color information in an image into a numerical vector, so that the AI ​​model can calculate and learn from these data; Learning module: Automatically extract key features from a large amount of cold source disaster-related data collected from the data acquisition module. For example, for image data taken by underwater cameras, the learning module can learn the shape, texture, color and other visual features of cold source disasters through technologies such as convolutional neural networks; for sonar data, it can learn the reflection intensity, echo time and other acoustic features of the target object, thereby converting these complex raw data into more valuable feature representations for identification and early warning.

Images