An AI-based intelligent chip bucket contents monitoring system and method

Abstract

This invention discloses an AI-based intelligent storage container content monitoring system and method, relating to the field of intelligent storage container content monitoring technology. It collects intelligent storage container data, extracts and preprocesses intelligent storage container features, divides the intelligent storage container's usage process into stages, and simultaneously stores the intelligent storage container's feature data. Furthermore, it analyzes the storage effect of the intelligent storage container's contents, comprehensively considering the intelligent storage container's usage process, analyzing the degree of influence of information features on each stage of use, and analyzing the process's impact. Based on the degree of influence of information features on each stage of use and the degree of influence of each stage of use on the storage effect of the usage process, it analyzes the process's impact, confirms the importance of information features to the usage process, performs anomaly detection in the intelligent storage container's usage process, and confirms the adjustment order of information features. This invention, through multi-dimensional data collection and refined feature analysis, ensures the stable quality of the contents throughout the entire process from filling to restoration.

Description

Technical Field

[0001] This invention relates to the field of intelligent bucket contents monitoring technology, specifically to an artificial intelligence-based intelligent bucket contents monitoring system and method. Background Technology

[0002] In smart storage applications, early monitoring technologies relied primarily on single-dimensional data collection, such as acquiring basic parameters like internal temperature and humidity through built-in sensors, lacking simultaneous tracking of related factors like the external environment and movement trajectory. This limitation prevented the comprehensive capture of dynamic changes during the storage process. Sudden changes in environmental temperature and humidity or disruptions during transport often failed to trigger timely warnings due to missing data. Furthermore, traditional methods lacked a scientific division of the usage process, relying solely on a crude "fill-storage-restore" model for evaluation. This made it difficult to accurately pinpoint the specific impact of each stage on the contents' condition, resulting in a lack of data support for tracing and improving storage quality issues. For example, contents might deviate from their intended state during transport due to environmental fluctuations, but the lack of feature annotation and analysis for this stage prevented targeted optimization of storage and transportation conditions.

[0003] Current technologies for anomaly detection and handling largely rely on preset thresholds or simple rules to trigger alarms, lacking in-depth data mining and intelligent learning. When internal stress values ​​exceed normal ranges, the system can only issue alarms but cannot quickly pinpoint the key factors causing the anomaly by combining experience from handling similar historical scenarios. Furthermore, anomaly adjustments often depend on manual intervention, requiring maintenance personnel to systematically investigate potential influencing factors, which is not only time-consuming and labor-intensive but also prone to errors due to subjective judgment biases. For example, when the contents are abnormal, manual adjustments may prioritize minor parameters while ignoring core features that determine storage performance, leading to recurring problems. This passive response model is ill-suited to complex and ever-changing usage scenarios, especially in large-scale applications, significantly increasing management costs and risks.

[0004] With the development of IoT technology, although some systems have introduced multi-source data acquisition, their data integration and analysis capabilities still have shortcomings. Environmental data, location data, and internal state data are often stored in independent databases, lacking a unified spatiotemporal benchmark and semantic association, leading to difficulties in cross-dimensional analysis. Meanwhile, existing models are mostly trained based on specific scenarios, lacking generalization ability and struggling to adapt to changes in the characteristics of different contents or environmental conditions. For example, a storage model designed for a certain type of liquid may fail when handling solid particulate contents due to incompatible feature extraction logic. Furthermore, the system lacks a dynamic optimization mechanism, unable to continuously learn from historical data to iterate standard processes and adjust strategies, resulting in a gradual decline in monitoring accuracy and response efficiency over time. This "static" management model is ill-suited to meet the demands of modern industry for refined and adaptive storage processes. Summary of the Invention

[0005] The purpose of this invention is to provide an artificial intelligence-based intelligent chip bucket contents monitoring system and method to solve the problems mentioned in the background art.

[0006] To address the aforementioned technical problems, this invention provides the following technical solution: a method for monitoring the contents of a smart bucket based on artificial intelligence, comprising the following steps:

[0007] S1. Collect data from the Smart Chip Bucket, and obtain the Smart Chip Bucket features after feature extraction and preprocessing.

[0008] S2. Divide the usage process of the Smart Chip Bucket into stages, synchronously store the characteristic data of the Smart Chip Bucket, and then analyze the storage effect of the contents of the Smart Chip Bucket.

[0009] S3. Taking into account the usage process of the Smart Chip Bucket, analyze the degree of impact of information characteristics on the usage stages. Information characteristics include internal data characteristics, predicted travel characteristics, and predicted environment characteristics.

[0010] S4. Analyze the degree of influence of the usage phase on the storage effect of the usage process. Based on the degree of influence of information characteristics on the usage phase and the degree of influence of the usage phase on the storage effect of the usage process, analyze the degree of influence of information characteristics on the storage effect of the usage process.

[0011] S5. Analyze the impact of the process, confirm the importance of information features to the usage process, and make anomaly judgments on the Smart Bucket usage process.

[0012] S6. When it is determined that the current Smart Bucket usage process is abnormal, confirm the adjustment order of information features, correct the information features in turn, until it is predicted that the current Smart Bucket process is normal.

[0013] Furthermore, in step S1, internal data of the smart bucket is collected through integrated sensors, and location data of the smart bucket is collected through the Beidou positioning system. Environmental data of the smart bucket's path is then obtained by combining this data with a meteorological platform. Feature processing is then performed on the internal data of the smart bucket to obtain its internal data features, including the internal temperature change rate and internal humidity change rate. Feature processing is also performed on the location data to obtain the predicted travel features of the smart bucket, including the ratio of the number of bumps to the preset number of bumps and the ratio of the number of impacts to the preset number of impacts. α observation points are set for the predicted path of the smart bucket, and feature extraction is performed on the environmental data of these observation points to obtain the predicted environmental features of the smart bucket, including the external temperature change rate and external humidity change rate. These internal data features, predicted travel features, and predicted environmental features are dimensionless. Through multi-dimensional data acquisition and feature processing, a comprehensive perception of the smart bucket's state is achieved. Internal data directly captures the core operating status, location information is transformed into predictable travel features, and environmental features extracted from meteorological resources allow for early detection of external changes along the path. These three types of information complement each other, avoiding the limitations of a single data source. The processed features accurately reflect key influencing factors, providing a solid foundation for subsequent stage division and effect analysis, ensuring the integrity and effectiveness of data from the source, and making the whole process monitoring more in line with the needs of actual scenarios.

[0014] Furthermore, in step S2, during the use of the Smart Chip Bucket with any type of content, the usage process of the Smart Chip Bucket is divided into stages, namely X usage stages. The usage process represents the set from the filling stage of filling the Smart Chip Bucket with content until the restoration stage of cleaning the Smart Chip Bucket. During the use of the Smart Chip Bucket, the internal data characteristics, predicted travel characteristics, and predicted environment characteristics of the Smart Chip Bucket are marked and stored in the Smart Chip Bucket usage log; the storage effect of the usage process is marked, which represents the consistency between the content before the filling stage and the content after the restoration stage.

[0015] For any given content, Y content features are defined. Feature collection is performed on the content before entering the filling stage and the content after completing the restoration stage. The Y content features of the content before entering the filling stage are {a1, a2, ..., a...} y ,…,a Y}, where a y This represents the y-th content feature of the content before the filling stage, and the y-th content features of the content after the restoration stage are {A1, A2, ..., A...}. y ,…,A Y}, where A y The y-th feature of the contents after the restoration phase is represented, and the uniformity between the contents before the filling phase and the contents after the restoration phase is denoted as W. A:

[0016] ;

[0017] By breaking down the usage process into stages, the complex process of using the Smart Bucket is decomposed into clear steps, facilitating precise identification of the impact of each stage on the contents and avoiding the ambiguity of a general analysis of the entire process. Simultaneously, internal, process, and environmental characteristics are labeled with corresponding stages and stored in logs, forming a complete process record chain, providing coherent data support for subsequent traceability and analysis. Furthermore, defining storage effectiveness with uniformity, and comparing the initial and final states of the contents, intuitively quantifies storage quality, ensuring the objectivity and specificity of the effectiveness evaluation, and laying a solid foundation for overall process optimization.

[0018] Furthermore, in step S3, the M usage processes in the historical data are analyzed, and the m-th usage process is analyzed, where m=1,2,…,M. The internal data characteristics of the Smart Chip Bucket are {C 1_m C 2_m ,…,C n1_m ,…,C N1_m}, the predicted travel feature is {D} 1_m D 2_m ,…,D n2_m ,…,D N2_m}, predicting environmental characteristics as {E 1_m E 2_m ,…,E n3_m ,…,E N3_m}, where N1 represents the number of internal data features, C n1_m D represents the n1th internal data feature, N2 represents the number of predicted travel features, and D n2_m N2 represents the n2th predicted trip feature, N3 represents the number of predicted environment features, and E n3_m Let {W} represent the storage effect of the n3rd predicted environmental feature and the Xth usage stage. 1_m W 2_m ,…,W x_m ,…,W X_m}, where W x_m This represents the storage effect of the x-th usage stage, where the storage effect of the x-th usage stage is the degree of consistency between the contents before entering the x-th usage stage and the contents after completing the x-th usage stage. The information feature in the m-th usage process is {F}. 1_m ,F 2_m ,…,F p_m ,…,F P_m}, where P represents the number of information features in the m-th usage process, and F p_mLet p represent the p-th information feature in the m-th usage process, and then calculate the degree of influence V of the p-th information feature on the x-th usage stage. p_x :

[0019] ;

[0020] Where f p_m Let p represent the p-th standard information feature in the m-th usage process. The method for obtaining the p-th standard information feature in the m-th usage process is as follows: statistically analyze the storage effect of the p-th stage of M usage processes, record the usage process with the largest storage effect as the standard usage process, record the information features in the standard usage process as standard information features, and thus obtain the p-th standard information feature in the m-th usage process; w x Let $\frac{p}{x}$ represent the average storage effect at the $x$-th stage in the $M$-th usage process. Substituting each value into $p = 1, 2, ..., P$, we obtain the degree of influence of $P$ information features on the $x$-th stage in the $m$-th usage process, $\frac{V}{\pi}$. 1_x V 2_x ,…,V p_x ,…,V P_x};

[0021] Through in-depth analysis of historical data, the impact of each information feature on the usage stage was quantified, enabling precise identification of factors influencing the process. By unifying the expression of three types of features and constructing a standard reference system, the one-sidedness of single-feature analysis was avoided, ensuring that the evaluation results are more aligned with actual scenarios. The clear definition of the impact at each stage allows the system to focus on key factors, providing targeted support for subsequent process optimization and reducing the randomness of adjustments. This data-driven analysis method improves the system's adaptability to complex environments and diverse content, enhances the reliability of predictions and the scientific nature of decision-making, and lays a solid foundation for achieving intelligent anomaly prediction.

[0022] Furthermore, in step S4, the degree of influence J of the storage effect of the x-th usage stage on the usage process is calculated. x :

[0023] ;

[0024] Where w m Let w represent the storage effect of the m-th usage process, and let w represent the average storage effect of the M usage processes.

[0025] The usage stages and information characteristics of M usage processes are uniformly defined, and the information characteristics are defined as {F1, F2, ..., F...} p ,…,F P}, and then obtain the information feature F p The degree of impact L on the storage effect of the usage processp :

[0026] ;

[0027] By substituting p = 1, 2, ..., P into the data, we obtain the degree of influence of P information features on the storage effect of the usage process. By quantifying the impact of each usage stage on the overall storage effect, we clearly define the weight of different stages in the entire process, avoiding a generalized analysis of the process's impact. Based on this, we further integrate the impact of information features on stages to form a direct impact assessment of features on the overall process, constructing a complete impact chain from features to stages to the entire process. This progressive analytical logic clarifies the global role of each information feature, highlighting the core position of key stages and accurately identifying features that play a decisive role in the overall storage effect. This provides a scientific basis for subsequent selection of effective features and judgment of process anomalies, improving the systematicity and accuracy of the entire process analysis.

[0028] Furthermore, in step S5, an impact threshold is set, and information features with an impact greater than the threshold are considered valid information features; otherwise, they are considered invalid information features. Q valid information features are collected in real time, and the deviation value of the predicted final storage effect is calculated. The deviation value of the final storage effect is the average of the product of the difference between the Q valid information features and the standard valid information feature and the impact level. The method for obtaining the standard valid information feature is as follows: the storage effect of M usage processes is statistically analyzed, and the usage process with the largest storage effect is recorded as the standard valid usage process, and the information features in the standard valid usage process are recorded as standard valid information features. When the deviation value of the final storage effect is less than or equal to the preset threshold value of the final storage effect, the current use of the Smart Bucket is predicted to be normal; otherwise, the current use of the Smart Bucket is predicted to be abnormal. By setting a threshold to filter valid information features, irrelevant factors are eliminated, allowing the analysis to focus on the core features that play a key role in the storage effect, thereby improving the efficiency and targeting of data processing. By using the features corresponding to optimal storage performance as the standard, and combining the deviation and impact of effective features to calculate the deviation value, anomaly judgment no longer relies on a single indicator, but takes into account both feature importance and actual deviation, thus enhancing the scientific rigor and accuracy of the judgment. By setting preset thresholds to clearly define whether the process is normal or not, it avoids the bias of subjective judgment and can complete anomaly early warnings in real time and efficiently, providing a reliable basis for timely intervention and adjustment, and ensuring the stability of the storage process.

[0029] In step S6, when the current Smart Bucket usage process is abnormal, the system calls the impact degree of Q valid information features and the difference between the Q valid information features and the standard valid information features. The priority adjustment coefficient of the q-th standard valid information feature is the product of the difference between the q-th valid information feature and the standard valid information feature and the impact degree, q=1,2,…,Q. Substituting q=1,2,…,Q one by one, the priority adjustment coefficients of the Q standard valid information features are obtained. The standard valid information features are sorted from largest to smallest according to the priority adjustment coefficient, and then the standard valid information features are adjusted one by one. After adjusting the standard valid information features once, the current Smart Bucket process is judged again to see if it is normal, until the current Smart Bucket process is predicted to be normal. When the process is abnormal, by comprehensively considering the impact degree of valid information features and the actual deviation, adjustments are made according to the criticality, avoiding the blindness of indiscriminate operation, and concentrating resources on the links with the most significant impact on storage effect, which greatly improves the efficiency of anomaly handling. After each adjustment, the process status is verified immediately, forming a closed loop of "adjustment-judgment" to ensure that each correction effectively moves closer to the normal state and gradually reduces deviations. This targeted and gradual optimization approach can quickly locate and solve core problems while reducing unnecessary operational interference, effectively enhancing the system's self-correction capabilities and ensuring that the Smart Bucket usage process can efficiently return to stability when anomalies occur.

[0030] An AI-based intelligent chip bucket contents monitoring system includes: a feature data processing module, a usage stage segmentation module, a feature stage impact module, a stage effect impact module, a process anomaly judgment module, and a feature adjustment and optimization module.

[0031] The feature data processing module is used to collect data from the smart chip bucket, and obtain the features of the smart chip bucket after feature extraction and preprocessing.

[0032] The usage phase segmentation module is used to segment the usage process of the Smart Chip Bucket into usage phases, synchronously store the characteristic data of the Smart Chip Bucket, and then analyze the storage effect of the contents of the Smart Chip Bucket.

[0033] The feature stage impact module is used to comprehensively consider the usage process of the Smart Bucket and analyze the degree of impact of information features on the usage stage.

[0034] The stage effect impact module is used to analyze the degree of impact of the usage stage on the storage effect of the usage process. Based on the degree of impact of information features on the stage and the degree of impact of the usage stage on the storage effect of the usage process, the module analyzes the degree of impact of information features on the storage effect of the usage process.

[0035] The process anomaly judgment module is used to analyze the degree of impact of the process, confirm the importance of information features to the process, and make anomaly judgments on the Smart Bucket usage process.

[0036] The feature adjustment and optimization module is used to determine the order of information feature adjustments when the current Smart Bucket usage process is abnormal, and to correct the information features in sequence until the current Smart Bucket process is predicted to be normal.

[0037] Compared with existing technologies, the beneficial effects achieved by this invention are as follows: On the one hand, through multi-dimensional data collection and refined feature analysis, it comprehensively captures key influencing factors during the use of the smart storage container, achieving full-process monitoring from internal state and movement trajectory to environmental changes. Combined with the scientific division of usage stages, it accurately correlates the characteristics of each stage with changes in the state of the contents, making the storage effect assessment more aligned with actual scenarios. This full-chain monitoring logic effectively avoids storage deviations caused by neglecting a single factor, significantly reducing the probability of contents deteriorating or becoming abnormal due to environmental or operational issues, and ensuring stable quality of the contents throughout the entire process from filling to restoration.

[0038] On the one hand, by relying on quantitative analysis of the impact of information features on the process, effective features that play a key role in storage performance can be quickly identified, directly addressing core deviations during anomaly detection. During the anomaly adjustment phase, by comprehensively ranking the impact level and feature differences, features with the greatest impact on the process are prioritized for correction, avoiding resource waste caused by blind adjustments. This targeted anomaly handling mechanism not only shortens problem-solving time but also reduces interference from invalid operations on the storage process, improving the fault tolerance and emergency response efficiency of the Smart Storage Bucket.

[0039] On the other hand, by leveraging standardized processes and impact assessment models built from historical data, the system can autonomously complete feature extraction, impact analysis, anomaly detection, and dynamic adjustments, reducing reliance on human experience. Through continuous learning and optimization of standard features and adjustment logic, this method can adapt to the usage needs of different contents and environments, achieving a shift from passive monitoring to proactive optimization. This intelligent management model not only enhances the ease of use of the Smart Storage Bucket but also provides reliable technical support for large-scale, diversified storage scenarios, helping storage processes develop towards greater efficiency and intelligence. Attached Figure Description

[0040] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0041] Figure 1 This is a structural diagram of an AI-based smart bucket contents monitoring system according to the present invention;

[0042] Figure 2 This is a flowchart of an artificial intelligence-based smart bucket content monitoring method according to the present invention. Detailed Implementation

[0043] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0044] Please see Figure 1 and Figure 2 This invention provides a technical solution: a method for monitoring the contents of a smart bucket based on artificial intelligence, comprising the following steps:

[0045] S1. Collect data from the Smart Chip Bucket, and obtain the Smart Chip Bucket features after feature extraction and preprocessing.

[0046] S2. Divide the usage process of the Smart Chip Bucket into stages, synchronously store the characteristic data of the Smart Chip Bucket, and then analyze the storage effect of the contents of the Smart Chip Bucket.

[0047] S3. Taking into account the usage process of the Smart Chip Bucket, analyze the degree of impact of information characteristics on the usage stages. Information characteristics include internal data characteristics, predicted travel characteristics, and predicted environment characteristics.

[0048] S4. Analyze the degree of influence of the usage phase on the storage effect of the usage process. Based on the degree of influence of information characteristics on the usage phase and the degree of influence of the usage phase on the storage effect of the usage process, analyze the degree of influence of information characteristics on the storage effect of the usage process.

[0049] S5. Analyze the impact of the process, confirm the importance of information features to the usage process, and make anomaly judgments on the Smart Bucket usage process.

[0050] S6. When it is determined that the current Smart Bucket usage process is abnormal, confirm the adjustment order of information features, correct the information features in turn, until it is predicted that the current Smart Bucket process is normal.

[0051] Furthermore, in step S1, data from inside the smart bucket is collected through integrated sensors, and location data is collected through the BeiDou positioning system. Environmental data along the smart bucket's path is then obtained by combining this data with a meteorological platform. Feature processing is then performed on the internal data of the smart bucket to obtain its internal data features. Feature processing is also performed on the location data to obtain the predicted travel features of the smart bucket. α observation points are set along the predicted path of the smart bucket, and feature extraction is performed on the environmental data of these observation points to obtain the predicted environmental features of the smart bucket. Through multi-dimensional data collection and feature processing, a comprehensive perception of the smart bucket's status is achieved. Internal data directly captures the core operating status, location information is transformed into predictable travel features, and environmental features extracted from meteorological resources allow for early detection of external changes along the path. These three types of information complement each other, avoiding the limitations of a single data source. The processed features accurately reflect key influencing factors, providing a solid foundation for subsequent stage division and effect analysis, ensuring the integrity and effectiveness of the data from the source, and making the entire process monitoring more aligned with actual scenario needs.

[0052] In step S2, the usage process of the Smart Chip Bucket is divided into stages for any type of content. The usage process is divided into X stages, which represent the set from the filling stage (filling the Smart Chip Bucket with content) to the restoration stage (cleaning the Smart Chip Bucket). During the use of the Smart Chip Bucket, the internal data characteristics, predicted travel characteristics, and predicted environment characteristics of the Smart Chip Bucket are labeled and stored in the Smart Chip Bucket usage log. The storage effect of the usage process is also labeled, which represents the consistency between the content before the filling stage and the content after the restoration stage.

[0053] For any given content, Y content features are defined. Feature collection is performed on the content before entering the filling stage and the content after completing the restoration stage. The Y content features of the content before entering the filling stage are {a1, a2, ..., a...} y ,…,a Y}, where a y This represents the y-th content feature of the content before the filling stage, and the y-th content features of the content after the restoration stage are {A1, A2, ..., A...}. y ,…,A Y}, where A y The y-th feature of the contents after the restoration phase is represented, and the uniformity between the contents before the filling phase and the contents after the restoration phase is denoted as W. A :

[0054] ;

[0055] By breaking down the usage process into stages, the complex process of using the Smart Bucket is decomposed into clear steps, facilitating precise identification of the impact of each stage on the contents and avoiding the ambiguity of a general analysis of the entire process. Simultaneously, internal, process, and environmental characteristics are labeled with corresponding stages and stored in logs, forming a complete process record chain, providing coherent data support for subsequent traceability and analysis. Furthermore, defining storage effectiveness with uniformity, and comparing the initial and final states of the contents, intuitively quantifies storage quality, ensuring the objectivity and specificity of the effectiveness evaluation, and laying a solid foundation for overall process optimization.

[0056] In step S3, the M usage processes in the historical data are analyzed, and the m-th usage process is analyzed, where m = 1, 2, ..., M. The internal data characteristics of the Smart Chip Bucket are {C}. 1_m C 2_m ,…,C n1_m ,…,C N1_m}, the predicted travel feature is {D} 1_m D 2_m ,…,D n2_m ,…,D N2_m}, predicting environmental characteristics as {E 1_m E 2_m ,…,E n3_m ,…,E N3_m}, where N1 represents the number of internal data features, C n1_m D represents the n1th internal data feature, N2 represents the number of predicted travel features, and D n2_m N2 represents the n2th predicted trip feature, N3 represents the number of predicted environment features, and E n3_m Let {W} represent the storage effect of the n3rd predicted environmental feature and the Xth usage stage. 1_m W 2_m ,…,W x_m ,…,W X_m}, where W x_m This represents the storage effect of the x-th usage stage, where the storage effect of the x-th usage stage is the degree of consistency between the contents before entering the x-th usage stage and the contents after completing the x-th usage stage. The information feature in the m-th usage process is {F}. 1_m ,F 2_m ,…,F p_m ,…,F P_m}, where P represents the number of information features in the m-th usage process, and F p_m Let p represent the p-th information feature in the m-th usage process, and then calculate the degree of influence V of the p-th information feature on the x-th usage stage. p_x :

[0057] ;

[0058] Where f p_m Let p represent the p-th standard information feature in the m-th usage process. The method for obtaining the p-th standard information feature in the m-th usage process is as follows: statistically analyze the storage effect of the p-th stage of M usage processes, record the usage process with the largest storage effect as the standard usage process, record the information features in the standard usage process as standard information features, and thus obtain the p-th standard information feature in the m-th usage process; w x Let $\frac{p}{x}$ represent the average storage effect at the $x$-th stage in the $M$-th usage process. Substituting each value into $p = 1, 2, ..., P$, we obtain the degree of influence of $P$ information features on the $x$-th stage in the $m$-th usage process, $\frac{V}{\pi}$. 1_x V 2_x ,…,V p_x ,…,V P_x};

[0059] Through in-depth analysis of historical data, the impact of each information feature on the usage stage was quantified, enabling precise identification of factors influencing the process. By unifying the expression of three types of features and constructing a standard reference system, the one-sidedness of single-feature analysis was avoided, ensuring that the evaluation results are more aligned with actual scenarios. The clear definition of the impact at each stage allows the system to focus on key factors, providing targeted support for subsequent process optimization and reducing the randomness of adjustments. This data-driven analysis method improves the system's adaptability to complex environments and diverse content, enhances the reliability of predictions and the scientific nature of decision-making, and lays a solid foundation for achieving intelligent anomaly prediction.

[0060] In step S4, the degree of influence J of the storage effect of the x-th usage stage on the usage process is calculated. x :

[0061] ;

[0062] Where w m Let w represent the storage effect of the m-th usage process, and let w represent the average storage effect of the M usage processes.

[0063] The usage stages and information characteristics of M usage processes are uniformly defined, and the information characteristics are defined as {F1, F2, ..., F...} p ,…,F P}, and then obtain the information feature F p The degree of impact L on the storage effect of the usage process p :

[0064] ;

[0065] By substituting p = 1, 2, ..., P into the data, we obtain the degree of influence of P information features on the storage effect of the usage process. By quantifying the impact of each usage stage on the overall storage effect, we clearly define the weight of different stages in the entire process, avoiding a generalized analysis of the process's impact. Based on this, we further integrate the impact of information features on stages to form a direct impact assessment of features on the overall process, constructing a complete impact chain from features to stages to the entire process. This progressive analytical logic clarifies the global role of each information feature, highlighting the core position of key stages and accurately identifying features that play a decisive role in the overall storage effect. This provides a scientific basis for subsequent selection of effective features and judgment of process anomalies, improving the systematicity and accuracy of the entire process analysis.

[0066] In step S5, an impact threshold is set, and information features with an impact greater than the threshold are considered valid information features; otherwise, they are considered invalid information features. Q valid information features are collected in real time, and the deviation value of the predicted final storage effect is calculated. The deviation value of the final storage effect is the average of the product of the difference between the Q valid information features and the standard valid information feature and the impact level. The method for obtaining the standard valid information feature is as follows: the storage effect of M usage processes is statistically analyzed, and the usage process with the largest storage effect is recorded as the standard valid usage process, and the information features in the standard valid usage process are recorded as standard valid information features. When the deviation value of the final storage effect is less than or equal to the preset threshold value of the final storage effect, the current use of the Smart Bucket is predicted to be normal; otherwise, the current use of the Smart Bucket is predicted to be abnormal. By setting a threshold to filter valid information features, irrelevant factors are eliminated, allowing the analysis to focus on the core features that play a key role in the storage effect, thereby improving the efficiency and targeting of data processing. By using the features corresponding to optimal storage performance as the standard, and combining the deviation and impact of effective features to calculate the deviation value, anomaly judgment no longer relies on a single indicator, but takes into account both feature importance and actual deviation, thus enhancing the scientific rigor and accuracy of the judgment. By setting preset thresholds to clearly define whether the process is normal or not, it avoids the bias of subjective judgment and can complete anomaly early warnings in real time and efficiently, providing a reliable basis for timely intervention and adjustment, and ensuring the stability of the storage process.

[0067] In step S6, when the current Smart Bucket usage process is abnormal, the system calls the impact degree of Q valid information features and the difference between the Q valid information features and the standard valid information features. The priority adjustment coefficient of the q-th standard valid information feature is the product of the difference between the q-th valid information feature and the standard valid information feature and the impact degree, q=1,2,…,Q. Substituting q=1,2,…,Q one by one, the priority adjustment coefficients of the Q standard valid information features are obtained. The standard valid information features are sorted from largest to smallest according to the priority adjustment coefficient, and then the standard valid information features are adjusted one by one. After adjusting the standard valid information features once, the current Smart Bucket process is judged again to see if it is normal, until the current Smart Bucket process is predicted to be normal. When the process is abnormal, by comprehensively considering the impact degree of valid information features and the actual deviation, adjustments are made according to the criticality, avoiding the blindness of indiscriminate operation, and concentrating resources on the links with the most significant impact on storage effect, which greatly improves the efficiency of anomaly handling. After each adjustment, the process status is verified immediately, forming a closed loop of "adjustment-judgment" to ensure that each correction effectively moves closer to the normal state and gradually reduces deviations. This targeted and gradual optimization approach can quickly locate and solve core problems while reducing unnecessary operational interference, effectively enhancing the system's self-correction capabilities and ensuring that the Smart Bucket usage process can efficiently return to stability when anomalies occur.

[0068] An AI-based intelligent chip bucket contents monitoring system, the system comprising: a feature data processing module, a usage stage segmentation module, a feature stage impact module, a stage effect impact module, a process anomaly judgment module, and a feature adjustment and optimization module;

[0069] The feature data processing module is used to collect data from the Smart Chip Bucket, and obtains the features of the Smart Chip Bucket after feature extraction and preprocessing.

[0070] The usage phase segmentation module is used to segment the usage process of the Smart Chip Bucket into usage phases, synchronously store the characteristic data of the Smart Chip Bucket, and then analyze the storage effect of the contents of the Smart Chip Bucket.

[0071] The Feature Stage Impact Module is used to comprehensively consider the usage process of the Smart Chip Bucket and analyze the degree of impact of information features on the usage stages.

[0072] The Phase Effect Impact Module is used to analyze the degree of impact of the usage phase on the storage effect of the usage process. It analyzes the degree of impact of information characteristics on the storage effect of the usage process based on the degree of impact of information characteristics on the phase of the usage phase and the degree of impact of the usage phase on the storage effect of the usage process.

[0073] The process anomaly detection module is used to analyze the degree of impact on the process, confirm the importance of information features to the process, and perform anomaly detection on the Smart Bucket usage process.

[0074] The feature adjustment and optimization module is used to determine the order of information feature adjustments when the current Smart Bucket usage process is abnormal, and to correct the information features in sequence until the current Smart Bucket process is predicted to be normal.

[0075] Example 1: During a single use of the Smart Bucket, the built-in sensors capture changes in internal temperature, humidity, and the state of its contents. A positioning system tracks its movement, and a weather platform provides information on environmental conditions such as temperature and wind speed along the transport route. This data is then processed: internal temperature and humidity fluctuations are transformed into characteristics reflecting the stability of the storage environment; the movement trajectory is converted into predictable travel patterns; and weather information along the route is transformed into environmental characteristics that may affect the contents. This comprehensive approach allows for a complete understanding of the Smart Bucket's real-time status.

[0076] The entire process of transporting fresh produce using the smart bucket is broken down into several clear stages, starting from the initial stage of loading fresh produce into the smart bucket, through intermediate transportation, temporary stops, unloading, and finally the final stage of cleaning the smart bucket. At each stage, the internal, travel, and environmental characteristics are recorded in the usage log. Simultaneously, the storage effectiveness of the entire process is measured by comparing the consistency in freshness and integrity between the initially loaded and finally unloaded fresh produce.

[0077] We analyzed data from past fresh produce transportation processes to assess the impact of different characteristics on each stage. For example, we selected the transportation process that historically maintained the best fresh produce condition as a reference and compared the temperature characteristics in the current process with those of the reference process to see how these differences affect the freshness of fresh produce during transportation. This allowed us to quantify the specific impact of temperature characteristics on each stage of transportation. The same method can also be applied to the analysis of other characteristics and stages.

[0078] The impact of each stage on the overall storage effect is analyzed; for example, the transportation stage may have a far greater impact on the final freshness of the produce than the temporary stopover stage. Based on this, the impact of each characteristic on the overall effect is further integrated; for example, temperature characteristics may have a greater impact on the freshness of the produce throughout the process than environmental wind characteristics.

[0079] Features that significantly impact storage performance, such as temperature and the degree of vibration during transportation, are identified. These key features are monitored in real time and compared with corresponding features in the historical best practices. The magnitude of the deviation is calculated, and then the degree of impact is considered to determine whether the current process is functioning correctly. If the deviation is within an acceptable range, the process is considered normal; otherwise, it is considered abnormal.

[0080] If the process is determined to be abnormal, such as a large deviation in temperature characteristics that significantly impacts storage performance, temperature-related parameters are adjusted first. Then, key characteristics are re-monitored to check if the process has returned to normal. If the abnormality persists, other characteristics are adjusted sequentially according to their degree of impact until the entire transportation process returns to normal.

[0081] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A method for monitoring the contents of a smart bucket based on artificial intelligence, characterized in that: The method includes the following steps: S1. Collect data from the Smart Chip Bucket, and obtain the Smart Chip Bucket features after feature extraction and preprocessing. S2. Divide the usage process of the Smart Chip Bucket into stages, synchronously store the characteristic data of the Smart Chip Bucket, and then analyze the storage effect of the contents of the Smart Chip Bucket. S3. Taking into account the usage process of the Smart Chip Bucket, analyze the degree of impact of information characteristics on the usage stages. Information characteristics include internal data characteristics, predicted travel characteristics, and predicted environment characteristics. S4. Analyze the degree of influence of the usage stage on the storage effect of the usage process. Based on the degree of influence of information characteristics on the usage stage and the degree of influence of the usage stage on the storage effect of the usage process, analyze the degree of influence of information characteristics on the storage effect of the usage process. S5. Analyze the impact of the process, confirm the importance of information features to the usage process, and make anomaly judgments on the Smart Bucket usage process. S6. When it is determined that the current Smart Bucket usage process is abnormal, confirm the adjustment order of information features, correct the information features in turn, until it is predicted that the current Smart Bucket process is normal. In step S1, data inside the smart bucket is collected by integrated sensors, location data of the smart bucket is collected by the Beidou positioning system, environmental data of the smart bucket's path is obtained by combining the meteorological platform, feature processing is performed on the data inside the smart bucket to obtain the internal data features of the smart bucket, feature processing is performed on the location data of the smart bucket to obtain the predicted travel features of the smart bucket, α observation points are set for the predicted path of the smart bucket, and feature extraction is performed on the environmental data of the observation points to obtain the predicted environmental features of the smart bucket; In step S2, the usage process of the Smart Chip Bucket is divided into stages for any type of content. The usage process is divided into X stages, which represent the set from the filling stage (filling the Smart Chip Bucket with content) to the restoration stage (cleaning the Smart Chip Bucket). During the use of the Smart Chip Bucket, the internal data characteristics, predicted travel characteristics, and predicted environment characteristics of the Smart Chip Bucket are labeled and stored in the Smart Chip Bucket usage log. The storage effect of the usage process is also labeled, which represents the consistency between the content before the filling stage and the content after the restoration stage.

2. The method for monitoring the contents of a smart bucket based on artificial intelligence according to claim 1, characterized in that: For any given content, Y content features are defined. Feature collection is performed on the content before entering the filling stage and the content after completing the restoration stage. The Y content features of the content before entering the filling stage are {a1, a2, ..., a...} y ,…,a Y }, where a y This represents the y-th content feature of the content before the filling stage, and the y-th content features of the content after the restoration stage are {A1, A2, ..., A...}. y ,…,A Y }, where A y The y-th feature of the contents after the restoration phase is represented, and the uniformity between the contents before the filling phase and the contents after the restoration phase is denoted as W. A : 。 3. The method for monitoring the contents of a smart bucket based on artificial intelligence according to claim 2, characterized in that: In step S3, the M usage processes in the historical data are analyzed, and the m-th usage process is analyzed, where m = 1, 2, ..., M. The internal data characteristics of the Smart Chip Bucket are {C}. 1_m C 2_m ,…,C n1_m ,…,C N1_m }, the predicted travel feature is {D 1_m D 2_m ,…,D n2_m ,…,D N2_m }, predicting environmental characteristics as {E 1_m E 2_m ,…,E n3_m ,…,E N3_m }, where N1 represents the number of internal data features, C n1_m D represents the n1th internal data feature, N2 represents the number of predicted travel features, and D n2_m N2 represents the n2th predicted trip feature, N3 represents the number of predicted environment features, and E n3_m Let {W} represent the storage effect of the n3rd predicted environmental feature and the Xth usage stage. 1_m W 2_m ,…,W x_m ,…,W X_m }, where W x_m This represents the storage effect of the x-th usage stage, where the storage effect of the x-th usage stage is the degree of consistency between the contents before entering the x-th usage stage and the contents after completing the x-th usage stage. The information feature in the m-th usage process is {F}. 1_m ,F 2_m ,…,F p_m ,…,F P_m }, where P represents the number of information features in the m-th usage process, and F p_m Let p represent the p-th information feature in the m-th usage process, and then calculate the degree of influence V of the p-th information feature on the x-th usage stage. p_x : ； Where f p_m The p-th standard information feature in the m-th usage process is represented by the following method: statistical analysis of the storage effect of the p-th stage of the M usage processes, recording the usage process with the largest storage effect as the standard usage process, recording the information features in the standard usage process as the standard information feature, and thus obtaining the p-th standard information feature in the m-th usage process. w x Let $\frac{p}{x}$ represent the average storage effect at the $x$-th stage in the $M$-th usage process. Substituting each value into $p = 1, 2, ..., P$, we obtain the degree of influence of $P$ information features on the $x$-th stage in the $m$-th usage process, $\frac{V}{\pi}$. 1_x V 2_x ,…,V p_x ,…,V P_x } 4. The method for monitoring the contents of a smart bucket based on artificial intelligence according to claim 3, characterized in that: In step S4, the degree of influence J of the storage effect of the x-th usage stage on the usage process is calculated. x : ； Where w m Let w represent the storage effect of the m-th usage process, and let w represent the average storage effect of the M usage processes. The usage stages and information characteristics of M usage processes are uniformly defined, and the information characteristics are defined as {F1, F2, ..., F...} p ,…,F P }, and then obtain the information feature F p The degree of impact L on the storage effect of the usage process p : ； Substitute each of the p=1,2,…,P to obtain the degree of influence of the P information features on the storage effect of the usage process.

5. The method for monitoring the contents of a smart bucket based on artificial intelligence according to claim 4, characterized in that: In step S5, an impact threshold is set, and information features with an impact greater than the threshold are considered valid information features; otherwise, they are considered invalid information features. Q valid information features are collected in real time, and the predicted final storage effect deviation value is calculated. The final storage effect deviation value is the average of the product of the difference between the Q valid information features and the standard valid information feature and the impact level. The standard valid information feature is obtained by: statistically analyzing the storage effect of M usage processes, recording the usage process with the largest storage effect as the standard valid usage process, and recording the information features in the standard valid usage process as standard valid information features. When the final storage effect deviation value is less than or equal to the preset final storage effect deviation value threshold, the current Smart Bucket usage process is predicted to be normal; otherwise, the current Smart Bucket usage is predicted to be abnormal.

6. The method for monitoring the contents of a smart bucket based on artificial intelligence according to claim 5, characterized in that: In step S6, when the current Smart Bucket usage process is abnormal, the system calls the impact degree of Q valid information features, and the difference between the Q valid information features and the standard valid information features. The priority adjustment coefficient of the q-th standard valid information feature is the product of the difference between the q-th valid information feature and the standard valid information feature and the impact degree, q=1,2,…,Q. Substituting q=1,2,…,Q one by one, the priority adjustment coefficients of the Q standard valid information features are obtained. The standard valid information features are sorted in descending order of priority adjustment coefficients, and then the standard valid information features are adjusted one by one. After adjusting the standard valid information features once, the current Smart Bucket process is judged again to see if it is normal, until the current Smart Bucket process is predicted to be normal.

7. An AI-based intelligent chip bucket content monitoring system, wherein the system is applied to the AI-based intelligent chip bucket content monitoring method according to any one of claims 1-6, characterized in that: The system It includes: a feature data processing module, a usage stage division module, a feature stage impact module, a stage effect impact module, a process anomaly judgment module, and a feature adjustment and optimization module; The feature data processing module is used to collect data from the smart chip bucket, and obtain the features of the smart chip bucket after feature extraction and preprocessing. The usage phase segmentation module is used to segment the usage process of the Smart Chip Bucket into usage phases, synchronously store the characteristic data of the Smart Chip Bucket, and then analyze the storage effect of the contents of the Smart Chip Bucket. The feature stage impact module is used to comprehensively consider the usage process of the Smart Bucket and analyze the degree of impact of information features on the usage stage. The stage effect impact module is used to analyze the degree of impact of the usage stage on the storage effect of the usage process. Based on the degree of impact of information features on the stage and the degree of impact of the usage stage on the storage effect of the usage process, the module analyzes the degree of impact of information features on the storage effect of the usage process. The process anomaly judgment module is used to analyze the degree of impact of the process, confirm the importance of information features to the process, and make anomaly judgments on the Smart Bucket usage process. The feature adjustment and optimization module is used to determine the order of information feature adjustments when the current Smart Bucket usage process is abnormal, and to correct the information features in sequence until the current Smart Bucket process is predicted to be normal.

Images