Cold-chain logistics multi-modal management method and system based on artificial intelligence

Abstract

The invention provides a cold-chain logistics multi-modal management method and system based on artificial intelligence, and the method comprises the steps: carrying out the dynamic monitoring and prediction of multi-modal environment parameters inside and outside a cold-chain container, and obtaining an environment parameter change prediction result within a preset time range; according to the environmental parameter change prediction result and the logistics route, the rotating speed of a compressor, the opening degree of an expansion valve and the refrigerant proportion of the refrigerating system are dynamically regulated and controlled, and a refrigerating strategy of the refrigerating system is obtained; based on a refrigeration strategy and a logistics route, the activation opportunity and proportion of the phase change material are calculated and controlled, and a temperature control scheme considering phase change point deviation under different altitude air pressures is obtained; and performing multi-mode dynamic management on the cold-chain logistics process according to the temperature control scheme and the refrigeration strategy. According to the method, complex environment changes in the multimodal transportation process can be effectively coped with, the refrigeration strategy and phase change material use are dynamically adjusted, and the quality and reliability of cold-chain logistics are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of servo drives, and in particular to an artificial intelligence-based multimodal management method and system for cold chain logistics. Background Art

[0002] In the existing cold chain logistics management technology, a single refrigeration system or a fixed temperature control strategy is usually used to maintain the temperature stability of the goods. However, this method is difficult to effectively cope with the complex environmental changes in the multimodal transport process, especially when switching between different transportation modes (such as from road transportation to air transportation), the drastic changes in environmental parameters (such as air pressure and temperature) will significantly affect the performance of the refrigeration system.

[0003] The main problem with existing technologies is that they cannot effectively adjust the refrigeration strategy and the use of phase change materials according to the dynamic changes of multimodal environmental parameters. This may lead to reduced refrigeration efficiency or unstable temperature control in certain special environments (such as plateau areas), affecting the quality and reliability of cold chain logistics. Summary of the invention

[0004] The main purpose of the present invention is to solve the technical problem that the existing cold chain logistics management method is difficult to effectively cope with complex environmental changes during multimodal transport; A first aspect of the present invention provides a multimodal management method for cold chain logistics based on artificial intelligence, and the multimodal management method for cold chain logistics based on artificial intelligence includes: Dynamically monitor and predict the multimodal environmental parameters inside and outside the cold chain container during the cold chain logistics process to obtain the prediction results of environmental parameter changes within a preset time range; According to the prediction results of the environmental parameter changes and the logistics route of the cold chain logistics, the compressor speed, expansion valve opening and refrigerant ratio of the refrigeration system are dynamically controlled to obtain a refrigeration strategy for the refrigeration system; Based on the refrigeration strategy and the logistics route, the activation timing and proportion of the phase change material in the cold chain container are calculated and controlled to obtain a temperature control solution that takes into account the phase change point offset under different altitude pressures; The cold chain logistics process is managed dynamically in a multi-modal manner according to the temperature control scheme and the refrigeration strategy.

[0005] Optionally, in a first implementation of the first aspect of the present invention, the dynamic monitoring and prediction processing of multimodal environmental parameters inside and outside the cold chain container during the cold chain logistics process to obtain the prediction results of environmental parameter changes within a preset time range includes: Perform multi-frequency sampling processing on the temperature, humidity, air pressure and light intensity inside and outside the cold chain container during the cold chain logistics process to obtain the original environmental parameter data set; Performing outlier detection and data cleaning processing on the original environmental parameter data set to obtain a preprocessed environmental parameter data set; Inputting the preprocessed environmental parameter data set into a preset multivariate time series prediction model to obtain preliminary environmental parameter prediction results; Using a preset integrated learning model and historical environmental data to perform prediction correction on the preliminary environmental parameter prediction results to obtain a corrected environmental parameter prediction result; The corrected environmental parameter prediction results are subjected to time series interpolation processing to obtain environmental parameter change prediction results within a preset time range.

[0006] Optionally, in a second implementation of the first aspect of the present invention, the compressor speed, expansion valve opening and refrigerant ratio of the refrigeration system are dynamically controlled according to the prediction result of the environmental parameter change and the logistics route of the cold chain logistics to obtain a refrigeration strategy for the refrigeration system, including: According to the prediction results of the environmental parameter changes, the expected temperature difference and the required refrigeration capacity in each transportation stage are calculated to obtain a refrigeration demand curve; Segmenting the logistics route, identifying key environmental change points, and obtaining a route feature sequence; Calculating the minimum power requirement of the compressor of the refrigeration system according to the refrigeration demand curve and the route characteristic sequence to obtain a basic speed curve of the compressor; Based on the compressor basic speed curve and the environmental parameter change prediction result, the compressor speed is dynamically adjusted to obtain a compressor speed control sequence; Calculating the optimal opening of the expansion valve according to the compressor speed control sequence and the required refrigeration capacity to obtain the expansion valve opening control sequence; Based on the prediction results of the environmental parameter changes, the ideal refrigerant state under different environments is calculated to obtain a refrigerant ratio adjustment plan; The compressor speed control sequence, the expansion valve opening control sequence and the refrigerant ratio adjustment scheme are time-synchronized to obtain a refrigeration strategy for the refrigeration system.

[0007] Optionally, in a third implementation of the first aspect of the present invention, calculating the minimum power requirement of the compressor of the refrigeration system according to the refrigeration demand curve and the route characteristic sequence to obtain the compressor basic speed curve includes: The refrigeration demand curve is subjected to piecewise linearization processing to obtain a piecewise refrigeration demand function, and the environmental characteristics of each transportation stage are quantitatively scored according to the route characteristic sequence to obtain an environmental complexity index sequence; The segmented refrigeration demand function is weightedly superimposed with the environmental complexity index sequence to obtain a comprehensive refrigeration load curve; According to the comprehensive refrigeration load curve, the theoretical minimum compression work is calculated using a thermodynamic formula to obtain a compression power curve, and the compression power curve is converted into a compressor speed to obtain a compressor basic speed curve.

[0008] Optionally, in a fourth implementation of the first aspect of the present invention, the activation timing and proportion of the phase change material in the cold chain container are calculated and controlled based on the refrigeration strategy and the logistics route to obtain a temperature control scheme that considers the phase change point offset at different altitudes and pressures, including: Extracting the altitude change sequence according to the logistics route, calculating the air pressure change curve during the cold chain logistics process, and obtaining the air pressure distribution in different transportation stages; Based on the air pressure distribution and the reference phase change temperature of the phase change material in the cold chain container, the phase change point offset under different air pressures is calculated to obtain a phase change point offset curve; Compare and analyze the refrigeration strategy with the phase change point offset curve to determine the activation timing and usage ratio of the phase change material and obtain a preliminary phase change material usage plan; According to the preliminary phase change material usage plan and refrigeration strategy, performance matching and optimization adjustments are carried out to obtain a temperature control plan that takes into account the phase change point offset under different altitude pressures.

[0009] Optionally, in a fifth implementation of the first aspect of the present invention, the phase change point offset under different air pressures is calculated based on the air pressure distribution and the reference phase change temperature of the phase change material in the cold chain container to obtain the phase change point offset curve, which includes: Discretizing the air pressure distribution to obtain a key air pressure point sequence, and establishing an air pressure-temperature relationship equation based on a reference phase change temperature of the phase change material to obtain a phase change point offset calculation model; For each air pressure value in the key air pressure point sequence, the phase change point offset calculation model is used to calculate to obtain a discrete phase change point offset value; The discrete phase change point offset values ​​are interpolated to obtain a continuous phase change point offset function, and a phase change point offset curve covering the entire air pressure range is generated according to the continuous phase change point offset function.

[0010] Optionally, in a sixth implementation of the first aspect of the present invention, the performing multimodal dynamic management of the logistics process of the cold chain logistics according to the temperature control scheme and the refrigeration strategy includes: According to the temperature control scheme and refrigeration strategy, the multimodal environmental parameters in the cold chain logistics process are monitored and analyzed in real time to obtain the current cold chain status evaluation results; Based on the cold chain status assessment result and in combination with preset management rules, control instructions for the refrigeration system and phase change materials are generated to obtain a resource allocation plan; The resource allocation scheme is converted into specific execution instructions, and the compressor speed, expansion valve opening, refrigerant ratio and activation of phase change materials of the refrigeration system are adjusted in real time to obtain a dynamic control instruction sequence; According to the dynamic control instruction sequence, the environmental parameters in the cold chain logistics process are continuously monitored and adjusted, and the regulation effect is recorded to form a closed-loop feedback control to achieve multi-modal dynamic management of cold chain logistics.

[0011] A second aspect of the present invention provides an artificial intelligence-based cold chain logistics multimodal management system, the artificial intelligence-based cold chain logistics multimodal management system comprising: The environmental prediction module is used to dynamically monitor and predict the multimodal environmental parameters inside and outside the cold chain container during the cold chain logistics process, and obtain the prediction results of environmental parameter changes within a preset time range; A refrigeration control module is used to dynamically control the compressor speed, expansion valve opening and refrigerant ratio of the refrigeration system according to the prediction results of the environmental parameter changes and the logistics route of the cold chain logistics, so as to obtain a refrigeration strategy for the refrigeration system; A phase change management module, for calculating and controlling the activation timing and proportion of the phase change material in the cold chain container based on the refrigeration strategy and the logistics route, and obtaining a temperature control solution that takes into account the phase change point offset under different altitude pressures; A dynamic management module is used to perform multimodal dynamic management of the cold chain logistics process according to the temperature control scheme and the refrigeration strategy.

[0012] The above-mentioned artificial intelligence-based cold chain logistics multimodal management method and system dynamically monitors and predicts the multimodal environmental parameters inside and outside the cold chain container to obtain the prediction results of environmental parameter changes within a preset time range; according to the prediction results of environmental parameter changes and the logistics route, the compressor speed, expansion valve opening and refrigerant ratio of the refrigeration system are dynamically adjusted to obtain the refrigeration strategy of the refrigeration system; based on the refrigeration strategy and logistics route, the activation timing and proportion of the phase change material are calculated and controlled to obtain a temperature control scheme that takes into account the phase change point offset under different altitude pressures; according to the temperature control scheme and refrigeration strategy, the cold chain logistics process is dynamically managed in a multimodal manner. The present invention can effectively cope with complex environmental changes in the multimodal transport process, dynamically adjust the refrigeration strategy and the use of phase change materials, and improve the quality and reliability of cold chain logistics.

[0013] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention are realized and obtained by the structures particularly pointed out in the description, claims and drawings.

[0014] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 Schematic diagram of a first embodiment of a multimodal management method for cold chain logistics based on artificial intelligence in an embodiment of the present invention; Figure 2 It is a schematic diagram of an embodiment of a cold chain logistics multimodal management system based on artificial intelligence in an embodiment of the present invention. DETAILED DESCRIPTION

[0016] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0017] The terms "including" and "having" and any variations thereof mentioned in the embodiments of the present invention are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device end including a series of steps or units is not limited to the listed steps or units, but may optionally include other steps or units that are not listed, or may optionally include other steps or units that are inherent to these processes, methods, products or device ends.

[0018] To facilitate understanding of this embodiment, firstly, a multimodal management method for cold chain logistics based on artificial intelligence disclosed in an embodiment of the present invention is introduced in detail. Figure 1 As shown, the method comprises the following steps: 101. Dynamically monitor and predict the multimodal environmental parameters inside and outside the cold chain container during the cold chain logistics process to obtain the prediction results of environmental parameter changes within a preset time range; In one embodiment of the present invention, the dynamic monitoring and predictive processing of multimodal environmental parameters inside and outside the cold chain container during the cold chain logistics process to obtain the prediction results of environmental parameter changes within a preset time range includes: multi-frequency sampling processing of the temperature, humidity, air pressure and light intensity inside and outside the cold chain container during the cold chain logistics process to obtain an original environmental parameter data set; outlier detection and data cleaning processing on the original environmental parameter data set to obtain a pre-processed environmental parameter data set; inputting the pre-processed environmental parameter data set into a preset multivariate time series prediction model to obtain preliminary environmental parameter prediction results; using a preset integrated learning model and historical environmental data to predict and correct the preliminary environmental parameter prediction results to obtain corrected environmental parameter prediction results; performing time series interpolation processing on the corrected environmental parameter prediction results to obtain environmental parameter change prediction results within a preset time range.

[0019] Specifically, in order to obtain sufficient precision of internal and external environmental parameters throughout the cold chain logistics transportation, multi-frequency sampling is required for temperature, humidity, air pressure and light intensity. In specific implementation, several high-precision sensors can be installed inside and outside the container, and they are synchronized with the data acquisition terminal at a unified time. Due to the different fluctuation periods of different parameters, the system will implement higher frequency sampling for light intensity that is susceptible to instantaneous interference, and adopt medium frequency for relatively stable indicators such as air pressure. The temperature and humidity will determine the appropriate collection cycle based on the actual temperature control requirements. After each sampling, the sensor will transmit the data to the central storage module via wired or wireless means, and attach metadata such as sampling time, geographic location and transportation stage. Under high altitude or extreme temperature conditions, the power supply and protection of the sensor will also be appropriately strengthened to ensure the stability of data collection. With such a multi-dimensional and variable resolution sampling strategy, the original environmental parameter data set can contain continuous information sufficient to reflect the dynamic changes of the internal and external environment of the container, and cover various complex scenarios experienced during the entire transportation process.

[0020] Specifically, after obtaining the above-mentioned original environmental parameter data set, outlier detection and data cleaning will be carried out immediately to ensure the accuracy and consistency of the data. The implementation method first analyzes the distribution of the measured values ​​in each time segment based on statistical methods. If a group of sensor readings obviously exceeds the credible range, it can be regarded as an outlier and marked or removed. For missing records caused by sensor failure or network interruption, they are usually filled with similar time values ​​of sensors of the same type or typical values ​​under the same historical conditions, while retaining the original missing position for subsequent traceability. If there is a serious inconsistency in the readings of multiple sensors at the same time, the normal operating sensor group will be referred to to determine which data has a greater degree of deviation, and the sensor calibration log will be compared to identify whether there is drift. During calibration, if there is a hardware failure log, the reading of the channel is directly excluded. If it is only a short-term fluctuation, it is revised by smoothing filtering or moving average method. After this process, the preprocessed environmental parameter data set can remove random noise and outliers while retaining the real environmental change trend as much as possible.

[0021] Specifically, after the preprocessing is completed, the system will send the cleaned environmental parameter data to the pre-built multivariate time series prediction model to obtain the preliminary environmental parameter prediction results for the future period. The model often uses a deep learning architecture, or a hybrid network combining LSTM and attention mechanism to extract the complex dependencies of temperature, humidity, air pressure and light intensity in the time dimension. When building the model, various environmental parameters are aligned at a uniform time interval, and the data of the previous several time steps are used as input features to infer the output of multiple future time steps. In the training phase, some historical data are usually deployed for cross-validation to determine the hyperparameters of the model and reduce overfitting. During operation, for the given latest time slice and its multidimensional parameter values ​​in the past several steps, the model will generate the values ​​of each predicted quantity moment by moment or segment by segment. In this way, the preliminary prediction results will relatively comprehensively reflect the dynamic coupling between multiple environmental factors, but there will still be certain errors due to short-term severe disturbances or special scenarios, which need to be corrected in subsequent steps to make the results closer to the substantial changes in the actual transportation process.

[0022] Specifically, after obtaining the preliminary prediction results, the system will call the integrated learning model to combine a large amount of historical environmental data to correct these prediction values, thereby reducing the deviation of a single model in special scenarios. The integrated learning model can include highly complementary sub-models such as random forests and various tree algorithms or shallow neural networks. The operation method is to compare the preliminary prediction results with the real sampling under similar conditions in historical data from multiple angles when capturing situations such as sudden changes in air pressure at high altitudes, huge temperature differences between day and night, or abnormal humidity jumps. If it is found that the preliminary prediction is too different from the past pattern, the correction rule will be activated to assign different weights to the output of each sub-model for comprehensive calculation to generate a value that is closer to the objective law. Seasonal changes or transportation route characteristics will also be considered during the correction process, and relevant records will be compared according to previously stored feature labels (such as geographic location, loading cargo attributes, etc.) to prevent overall judgment distortion due to local anomalies. The final corrected environmental parameter prediction results will be better than the preliminary prediction in terms of accuracy and stability, and a more accurate quantitative evaluation of the coupling between multiple variables can also be made.

[0023] Specifically, time series interpolation processing is performed on the basis of the corrected environmental parameter prediction results, which will form a smoother and continuous numerical distribution within the preset time range and avoid abrupt jumps between prediction points. In specific implementation, piecewise linear interpolation or high-order interpolation based on spline functions and other methods will be used according to the required output time resolution. The first step is usually to use the discrete moment where the predicted value is located as the key node, and calibrate and connect the values ​​of each parameter along the time axis; the second step is to calculate the environmental parameter value at each intermediate moment between the nodes according to the selected interpolation method. The piecewise linear method directly connects adjacent predicted values ​​to generate a concise and stable transition curve; spline interpolation uses continuous and high-order differentiable polynomial functions to make the curve smoother between nodes. This can keep the temperature, humidity, air pressure and light intensity better coherent on a high-resolution time scale and better fit the gradual characteristics of the actual physical process. The final interpolated sequence will be stored in the prediction result library and the corresponding error range will be marked.

[0024] 102. According to the prediction results of environmental parameter changes and the logistics route of cold chain logistics, the compressor speed, expansion valve opening and refrigerant ratio of the refrigeration system are dynamically controlled to obtain the refrigeration strategy of the refrigeration system; In one embodiment of the present invention, the compressor speed, expansion valve opening and refrigerant ratio of the refrigeration system are dynamically controlled according to the environmental parameter change prediction results and the logistics route of the cold chain logistics to obtain a refrigeration strategy for the refrigeration system, including: calculating the expected temperature difference and required refrigeration capacity of each transportation stage according to the environmental parameter change prediction results to obtain a refrigeration demand curve; segmenting the logistics route to identify key environmental change points to obtain a route feature sequence; calculating the minimum power requirement of the compressor of the refrigeration system according to the refrigeration demand curve and the route feature sequence to obtain a compressor basic speed curve; dynamically adjusting the compressor speed based on the compressor basic speed curve and the environmental parameter change prediction results to obtain a compressor speed control sequence; calculating the optimal opening of the expansion valve according to the compressor speed control sequence and the required cooling capacity to obtain an expansion valve opening control sequence; calculating the ideal refrigerant state under different environments based on the environmental parameter change prediction results to obtain a refrigerant ratio adjustment plan; performing time synchronization processing on the compressor speed control sequence, the expansion valve opening control sequence and the refrigerant ratio adjustment plan to obtain a refrigeration strategy for the refrigeration system.

[0025] Specifically, according to the predicted results of the environmental parameter changes, the expected temperature difference and the required refrigeration capacity of each transportation stage are calculated to obtain the implementation process of the refrigeration demand curve. The external temperature of each time slice, the target temperature inside the container, and the heat loss of the transport load can be extracted from the known environmental prediction data. The statistics of the external temperature can be combined with the diurnal changes and regional climate characteristics to divide the temperature curve into multiple time periods, each of which corresponds to a relatively stable environmental condition. When the temperature difference and the heat dissipation related to the characteristics of the goods are included in the calculation of the refrigeration capacity, it is necessary to conduct a thermal analysis of the process of heat dissipation or absorption of the goods, including respiration, the influence of relative humidity on heat exchange efficiency, etc. When calculating, the temperature difference of each time period is usually multiplied by the estimated heat flow per unit time to obtain the refrigeration demand value of the time period. Then, these refrigeration capacities are summarized point by point on the time axis of the entire transportation cycle to form a refrigeration demand curve that changes with time. For the instantaneous period when the temperature difference rises or falls sharply, a segmented weighting process will be performed to ensure that the peak demand is accurately presented during the period of sudden high or low temperature. Then, the results of each segment are interpolated and smoothed to output a refrigeration demand curve that can be used continuously. This curve usually has a unified time scale and is synchronized with the internal clock referenced by subsequent control links. It is used to guide the subsequent quantitative control of adjustment objects such as compressor speed, thereby fully reflecting the changes in demand in each transportation stage and reflecting the differentiated impact of environmental forecast data on different time periods.

[0026] Specifically, the process of segmenting the logistics route, identifying key environmental change points, and obtaining the route feature sequence requires dividing the intervals according to the change nodes of different altitudes, climate zones or docking links in the transportation path. First, it is necessary to locate each major turning point of the route on the geographic information system, extract the corresponding altitude and historical climate type, and pair these locations with the time axis. Then, find the moments of large altitude differences or regional climate type changes in the time series, which are often accompanied by obvious sudden changes in temperature, air pressure or humidity, and also include environmental changes caused by switching from sea to road or road to air. Aggregate and analyze the weather data and predicted events (such as customs inspections or nighttime shutdowns) of each section, and mark the start and end times, section lengths, and environmental trends as feature tags. When establishing the route feature sequence, these section information will be stored in a data structure in a sequential manner, and each section will be assigned a separate identifier and a retrievable description of geography, meteorology, and transportation methods. The route feature sequence established in this way can clearly define the specific external environmental characteristics and duration of each stage, so that the compressor speed and refrigerant ratio can be combined with the section attributes rather than blindly adjusted according to time, thereby achieving more targeted dynamic control.

[0027] Specifically, when calculating the minimum power demand of the compressor of the refrigeration system and obtaining the basic speed curve of the compressor according to the refrigeration demand curve and the route characteristic sequence, it is necessary to integrate the coupling effect of the thermodynamic model and the environmental factors on the refrigeration load. When implementing, the refrigeration capacity of each period in the refrigeration demand curve is usually read first, and the refrigeration capacity of the high altitude section or the extremely high temperature section is corrected by an additional coefficient in combination with the environmental complexity information of the corresponding route characteristic sequence. The corrected refrigeration load curve can be regarded as a continuously changing refrigeration demand target, and the minimum power required per unit time of the compressor is calculated by thermodynamic formulas and equipment performance parameters. During the calculation process, it is necessary to refer to the energy efficiency ratio of the compressor under different inlet pressures and refrigerant flow rates, as well as the limit constraints on the operation of the unit when the external environment has short-term strong fluctuations. When these comprehensive results are converted into speed parameters, an approximate function between speed and power is generally preset or the speed-power mapping relationship is obtained through equipment calibration data, so that each period can match the basic speed that meets the minimum refrigeration demand and meets the safety of compressor operation. The final compressor basic speed curve will cover the entire transportation cycle, and a speed change trajectory that coincides with the fluctuation of refrigeration demand will appear in the critical period. The curve usually takes time or segment index as the horizontal axis and compressor speed as the vertical axis, and is used to guide further dynamic control adjustments.

[0028] Specifically, when dynamically adjusting the compressor speed to obtain the compressor speed control sequence based on the compressor basic speed curve and the predicted results of environmental parameter changes, it is necessary to additionally monitor the real-time or near-real-time external environmental fluctuations. During implementation, the difference between the current compressor speed and the reference value on the basic curve is retrieved at each time slice or when a key event is triggered. If the environmental parameters suddenly deviate from the predicted range, such as high temperature or strong sunlight causing the load to exceed the basic estimated value, the speed will be corrected upward; if the external temperature drops or the heat dissipation of the cargo decreases, the speed will also be adjusted downward to reduce excessive consumption. In order to control the frequency of starting and stopping the compressor, the algorithm will set the adjustment threshold and consider the mechanical inertia and equipment life when the speed changes to ensure that there will be no large fluctuations in a short period of time. At the same time, the latest updated value of the multivariate time series prediction will be compared. If the prediction results show that there will be extreme conditions in a certain future period, a buffering adjustment will be made in advance at the current moment to make the speed change smoother. The speed control sequence thus formed contains the command speed value at each time node, which can achieve flexible response and balance energy consumption and refrigeration effect. This sequence is often recorded in the control scheduling system with finer time steps, and clear speed setting instructions are issued to the execution unit.

[0029] Specifically, when calculating the optimal opening of the expansion valve and obtaining the expansion valve opening control sequence according to the compressor speed control sequence and the required cooling capacity, it is necessary to pay attention to the evaporator pressure, actual refrigerant flow rate and superheat. When implementing, the current speed and cooling capacity setting value can be combined with the refrigerant circulation volume in the system pipeline to determine at what opening a stable low temperature zone can be maintained at the evaporation end. The temperature and pressure at the outlet of the evaporator are monitored by sensors to analyze whether the superheat is within a safe range; if the superheat is too high, it means that the expansion valve opening is too small and the refrigerant flow rate is insufficient, and the opening needs to be appropriately increased; if the superheat is too low, it means that the refrigerant is not fully vaporized or absorbs heat, and the opening needs to be reduced to stabilize the operation. In the long term, the system will subdivide the expansion valve opening according to the changing trend of the aforementioned refrigeration demand, so that the amount of refrigerant entering the evaporator can match the dynamic load in time. In the short-term mutation stage, the expansion valve will also be coordinated and linked according to the instantaneous speed increase or decrease of the compressor to avoid insufficient refrigerant supply at high load or energy waste at low load. The expansion valve opening control sequence finally formed by this process will be marked in the timing table of the control system together with the speed command, and will be appropriately delayed or synchronized according to the execution response speed of the specific equipment to ensure that the overall refrigeration cycle remains effectively connected.

[0030] Specifically, when calculating the ideal refrigerant state under different environments and obtaining the refrigerant ratio adjustment plan based on the prediction results of environmental parameter changes, it is necessary to analyze the characteristics of multiple refrigerants that may be used in the system, and compare the saturated pressure and condensation temperature of the refrigerant in combination with the dynamic changes of external temperature, air pressure and humidity. During implementation, several refrigerants with large differences in physical properties are often selected for mixing, and the heat transfer efficiency and compression load under different environmental conditions are optimized by changing the mixing ratio. If the transportation route includes a low-temperature and low-pressure stage, it is necessary to increase the content of a certain refrigerant with a lower evaporation temperature at this stage; if the route has high-temperature and high-humidity conditions, it is necessary to increase the content of refrigerants that can maintain good condensation efficiency in the high-temperature zone. By reading the external temperature and air pressure values ​​output by the environmental prediction model at each moment, the current ratio basis of the refrigeration system will be given priority when calculating the ratio adjustment strategy, and the feasibility of switching or progressive allocation will be evaluated. If the ratio needs to be changed frequently, a phased transition can be made in combination with the design of the refrigerant storage tank and pipeline switching to avoid large fluctuations. The final ratio adjustment scheme can be expressed as a ratio matrix that changes over time or segment, with each time period having a reasonable mixing ratio to achieve better cooling performance and energy consumption performance under changing environments.

[0031] Specifically, when the compressor speed control sequence, the expansion valve opening control sequence and the refrigerant ratio adjustment scheme are time-synchronized and the refrigeration strategy of the refrigeration system is obtained, it is necessary to comprehensively consider the physical delays and interactive effects of each adjustment object during the execution process. During implementation, the above three sequences are usually mapped to a unified time axis, and the target values ​​of each sequence at different times are compared. If the expansion valve opening needs to be increased suddenly at a certain moment, the system should confirm that the compressor speed command has been increased accordingly at a slightly earlier time, otherwise it may cause a sudden change in the evaporation pressure. At the same time, when switching the refrigerant ratio, it is necessary to ensure that the speed and opening maintain a reasonable balance before the ratio adjustment is completed to prevent unnecessary overshoot or lag caused by increasing or decreasing the refrigerant flow before the mixing ratio is stable. The process of time synchronization processing includes interpolation alignment and sequence verification, accurately aligning all instruction nodes to the same time slice, and then performing a gradual transition according to the equipment response lag to ensure that there will be no obvious impact inside the system. After synchronization is completed, a cooling strategy schedule covering all control elements will be output, including compressor speed settings, expansion valve opening parameters and refrigerant ratios for different time periods, for the controller to execute according to the time step. This integrated strategy enables the system to have control measures that match the environment and load at each transportation stage.

[0032] Furthermore, the method of calculating the minimum power requirement of the compressor of the refrigeration system according to the refrigeration demand curve and the route characteristic sequence to obtain the compressor basic speed curve includes: performing piecewise linearization processing on the refrigeration demand curve to obtain a piecewise refrigeration demand function, and quantitatively scoring the environmental characteristics of each transportation stage according to the route characteristic sequence to obtain an environmental complexity index sequence; weightedly superimposing the piecewise refrigeration demand function and the environmental complexity index sequence to obtain a comprehensive refrigeration load curve; calculating the theoretical minimum compression work according to the comprehensive refrigeration load curve using thermodynamic formulas to obtain a compression power curve, and converting the compression power curve into a compressor speed to obtain a compressor basic speed curve.

[0033] Specifically, the cooling demand curve is processed piecewise linearly to obtain a piecewise cooling demand function, and the environmental characteristics of each transportation stage are quantitatively scored according to the route characteristic sequence to obtain the environmental complexity index sequence. The cooling demand curve of the entire transportation cycle can be divided into several segments according to time or segment, and a linear fit is performed on the demand interval that shows a relatively stable or approximately linear change in each segment. The interval is , the linear function can be expressed as ; in and is the coefficient obtained by fitting by least squares or other methods. The smaller the mean square error of the segmented result in each interval, the higher the approximation of the linear function to the actual demand. Afterwards, the environmental factors in the transportation path should be investigated in combination with the route feature sequence. For example, when the altitude rises suddenly, resulting in drastic fluctuations in air pressure or an increase in the temperature difference between day and night, each stage should be scored to form an environmental complexity index sequence. This index can add parameters such as extreme temperature differences and sudden changes in humidity into a single value. For example, a higher complexity score is given to sections with high heat loads and significant air pressure fluctuations, and a lower score is given to sections with gentle environmental changes. These scores are arranged into vectors according to the time dimension so that each transportation stage has a clear complexity index, which is matched with the segmented demand function at the corresponding time. The segmented refrigeration demand function is weightedly superimposed with the environmental complexity index sequence to obtain the comprehensive refrigeration load curve. It is necessary to multiply the linear demand value at each moment by the environmental complexity index at the corresponding moment, or use a more flexible function operation to reflect the amplifying effect of environmental challenges on the refrigeration load. If the segmented demand function is set in the interval The value inside is , the value of the environmental complexity index at this moment is , then the comprehensive cooling load can be defined as ; in and is the coefficient used to control the weight of basic demand and environmental increment. In this formula, represents the general demand part, and It represents the additional load caused by environmental complexity. If a period of time has high temperature, high humidity or drastic altitude fluctuations, then The value is large, and the final calculated After completing the calculations in each interval, the system load can be increased. Splicing and interpolation processing are performed to form a continuous comprehensive cooling load curve on the time axis and mark the specific moments of load peaks and valleys to facilitate the next step of thermodynamic power calculation.

[0034] Specifically, according to the comprehensive refrigeration load curve, the theoretical minimum compression work is calculated using thermodynamic formulas to obtain the compression power curve. The load curve needs to be regarded as the time distribution of the heat absorbed by the evaporator side of the system, and is deduced according to the quantitative relationship between the gas compression and heat release process in the compressor working cycle. If the ideal gas model is used and based on the reversible adiabatic compression setting, the compressor can be set at time The intake pressure is , the exhaust pressure is , the specific heat ratio is , then the ideal compression work per unit time can be expressed as in is the refrigerant mass flow rate through the compressor, is the gas constant, is the evaporation temperature, is the isentropic efficiency of the compressor. If the comprehensive refrigeration load curve is at time Mapping to the evaporator heat absorption, it can be indirectly derived and Changes. Numerical integration or point-by-point discrete accumulation can be performed for the entire transportation period to obtain the theoretical minimum power value at each discrete moment to form a compression power curve. For sections where temperature and pressure change rapidly, correction items can be added according to the actual lag and throttling loss factors. After the calculation is completed, the compressor power demand at each moment can be obtained and presented on the time axis to observe the energy consumption trend of the load at different time periods such as high altitude and day and night, and convert the compression power curve into the compressor speed. When obtaining the compressor basic speed curve, it is necessary to interpolate or approximate the relationship between power and speed based on the known compressor's own performance curve. The shaft power and speed of most positive displacement compressors are nonlinearly related. A mapping function can be established using the calibration data provided by the equipment manufacturer or the experimental calibration results. For example, let the speed is the independent variable, power As the dependent variable, a set of discrete sampling points is formed , and then polynomial fitting or piecewise interpolation method is used to obtain After obtaining the target power value at each moment, if the inversion mapping function , we can solve , thereby generating a basic speed curve on the time axis. If there is a physical limit on the speed in the high-speed operation section or at low load, the curve should be truncated or limited according to the safety threshold to prevent the compressor from exceeding the design range at extreme speeds. The generated basic speed curve is regarded as a reference for subsequent dynamic control. It can also be used to evaluate the switching timing between segmented low-energy consumption mode and full-load mode in special circumstances, and to analyze the risk of frequent start and stop of the compressor. The final curve will match other elements in the environmental parameter prediction to support more sophisticated automatic scheduling logic.

[0035] 103. Based on the refrigeration strategy and logistics route, the activation timing and proportion of the phase change material in the cold chain container are calculated and controlled to obtain a temperature control solution that takes into account the phase change point offset under different altitude pressures; In one embodiment of the present invention, the activation timing and proportion of the phase change material in the cold chain container are calculated and controlled based on the refrigeration strategy and the logistics route to obtain a temperature control scheme that takes into account the phase change point offset under different altitude pressures, including: extracting the altitude change sequence according to the logistics route, calculating the air pressure change curve in the cold chain logistics process, and obtaining the air pressure distribution in different transportation stages; based on the air pressure distribution and the reference phase change temperature of the phase change material in the cold chain container, calculating the phase change point offset under different air pressures to obtain a phase change point offset curve; comparing and analyzing the refrigeration strategy with the phase change point offset curve to determine the activation timing and usage ratio of the phase change material, and obtain a preliminary phase change material usage plan; according to the preliminary phase change material usage plan and the refrigeration strategy, performance matching and optimization adjustment are performed to obtain a temperature control scheme that takes into account the phase change point offset under different altitude pressures.

[0036] Specifically, the altitude change sequence is extracted according to the logistics route, the pressure change curve in the cold chain logistics process is calculated, and the implementation process of obtaining the pressure distribution in different transportation stages can first determine the geographical coordinates and corresponding altitude information of each key node in the transportation route planning file, and arrange these nodes into a continuous spatial path according to the actual driving order. Then, the altitude in each section is sampled in detail, and the sampling interval can be dynamically set according to the line length and change rate to ensure that sufficiently dense data is collected in sections with large altitude fluctuations. If there is a connection by air or rail, it is also necessary to additionally mark the height difference in the flight section or station section, and connect the coordinates with the highway section. After completing the altitude collection, these altitude values ​​are combined with known meteorological models or empirical formulas to convert the corresponding air pressure values, such as calculating the average air pressure at each sampling point based on the exponential decay law of standard atmospheric pressure with altitude or a multi-maintenance correction model. If the altitude changes in some sections during transportation are extremely steep, the resolution can be increased during the interpolation process to capture significant pressure fluctuations within a short distance, and a small amount of outlier data can be eliminated or smoothed. Finally, according to the vehicle speed, stop time and altitude difference in different periods or sections of the entire route, a complete air pressure change curve is output to identify the atmospheric pressure conditions in each transportation stage and distinguish plateaus, low-altitude plains and transition areas on the curve.

[0037] Specifically, based on the air pressure distribution and the reference phase change temperature of the phase change material in the cold chain container, the phase change point offset under different air pressures is calculated to obtain the phase change point offset curve. The phase change temperature of the phase change material under standard sea level atmospheric pressure is used as the reference value, and the numerical calculation is performed in combination with the thermodynamic experimental data or the air pressure-phase change temperature correction function provided by the manufacturer. If the correction function is in the form of ,but It can be interpolated or fitted with the air pressure A nonlinear mapping relationship is established. Then, the pressure curve values ​​obtained in the previous step are substituted point by point in the time series or road segment index. , calculate the phase change temperature offset at each sampling point one by one, and add or subtract the result from the reference temperature to generate a phase change point curve that changes with altitude or transportation stage. If there is a steep phase change temperature jump in certain pressure intervals, it is necessary to discretize the curve more finely in this interval to avoid triggering or delaying the cold storage and cooling before the material enters the expected phase change state in actual operation. After the discrete calculation is completed, the data points can be uniformly interpolated to obtain a continuous and smooth phase change point offset curve, so that the actual transition temperature of the material can be queried at any given time or geographical location.

[0038] Specifically, the refrigeration strategy is compared and analyzed with the phase change point offset curve to determine the activation timing and usage ratio of the phase change material. The process of obtaining the preliminary phase change material usage plan requires point-by-point comparison of the operation plan of the refrigeration system in different temperature ranges with the phase change temperature of the material under the corresponding air pressure. First, read the target cabin temperature, compressor load curve and the cooling capacity that can be provided in each period of the refrigeration strategy, and identify the time or section that requires additional cold storage support or the need to reduce the cooling capacity. Then, according to the temperature threshold in the phase change point offset curve, find out which period of time the ambient pressure and cabin heat exchange conditions enable the material to successfully complete heat absorption or heat release. If the predicted cabin temperature is below the phase change temperature of the material, the heat release process will not be triggered. If it is higher than the phase change temperature, the material will enter the cooling mode. At this time, the activation time of the material can be evaluated according to the freshness demand and cooling margin of the goods, and it can be determined to open or gradually open some phase change materials within a certain period of time to absorb excess heat; or close the phase change process of some materials under low temperature conditions at night, so as to keep the refrigeration capacity in the optimal range. The proportion of material usage can refer to the overall volume and cargo layout of the cold chain container. If the direct contact area between the material and the cargo is too large, the heat released or absorbed during the phase change will cause a local temperature gradient, and the temperature distribution needs to be more uniform by adjusting the number of material modules or the layering method. After this analysis, a preliminary usage plan can be output, recording the number and duration of phase change materials activated in each period of time, and marking the temperature threshold and startup delay connected with the refrigeration strategy.

[0039] Specifically, according to the preliminary phase change material use plan and refrigeration strategy, the performance matching and optimization adjustment are carried out to obtain the temperature control plan that takes into account the phase change point offset at different altitudes. The process needs to include the real heat storage and release process of the material into the system simulation or simulation plus experimental verification to evaluate whether the phase change material can maintain a coordinated state with the compressor, expansion valve and refrigerant ratio at a specific altitude and temperature. If the simulation shows that under certain high-altitude working conditions, the phase change material is delayed in use due to excessive phase change temperature offset, then temperature control deviation will occur during low load or high load shock. At this time, it is necessary to adjust the activation sequence of the material or reduce the amount of material input at a single time to avoid overreaction, and it is also necessary to link with the start-stop frequency and refrigerant flow of the refrigeration system. If the material releases heat too quickly in some sections, causing a short-term jump in the cabin temperature, the operation of segmented pre-cooling or fine-tuning the phase change temperature threshold can be inserted into the strategy to allow the material to slowly release heat. When various corrections and simulation results meet the boundary conditions of cargo preservation and equipment safety, these tuned parameters can be solidified into the final plan, and the plan can be archived according to the transportation stage, cabin temperature control range and external environmental pressure. The temperature control plan generated in this way can be sent to the execution control system before departure or in real-time operation, so that it can flexibly control the timing and amount of use of phase change materials when facing different altitudes and pressures, ensuring that the cabin temperature remains in a safe range while taking into account energy efficiency, thus forming a complete temperature management system that includes phase change point offset correction content.

[0040] Furthermore, the phase change point offset under different air pressures is calculated based on the air pressure distribution and the reference phase change temperature of the phase change material in the cold chain container to obtain the phase change point offset curve, including: discretizing the air pressure distribution to obtain a key air pressure point sequence, and establishing an air pressure-temperature relationship equation according to the reference phase change temperature of the phase change material to obtain a phase change point offset calculation model; calculating each air pressure value in the key air pressure point sequence using the phase change point offset calculation model to obtain a discrete phase change point offset value; interpolating the discrete phase change point offset values ​​to obtain a continuous phase change point offset function, and generating a phase change point offset curve covering the entire air pressure range based on the continuous phase change point offset function.

[0041] Specifically, the air pressure distribution is discretized to obtain a sequence of key air pressure points, and the air pressure-temperature relationship equation is established according to the reference phase change temperature of the phase change material to obtain the implementation process of the phase change point offset calculation model. The air pressure distribution data at different stages of the entire cold chain logistics route or aviation, plateau transportation, etc. will be collected first, and this distribution will be split according to a certain time step or space step. If the original air pressure curve presents a more complex multi-peak or multi-segment trend, it is necessary to set a discrete threshold and sampling interval according to its fluctuation amplitude and frequency, so that each key air pressure point can represent the typical air pressure level in the section. In order to prevent the neglect of sampling in certain air pressure intervals, the sampling density will be increased in the intervals with more drastic changes in air pressure, and the number of samplings will be moderately reduced in relatively stable sections, thereby avoiding the processing load caused by excessive data. The screened air pressure points are arranged in order from low to high or from high to low to form a sequence of key air pressure points and the position and time of each point are marked, so that the sequence can be quickly located to the corresponding geographical location and logistics stage through indexing. Then, the reference phase change temperature of the phase change material under standard atmospheric pressure is retrieved, and the existing pressure-temperature correction rules are used to construct a more widely adaptable pressure-temperature relationship equation. If the phase change process of the phase change material conforms to certain thermodynamic approximations, the improved Clausius-Clapeyron equation or polynomial fitting method can be selected to describe the relationship between the melting point or freezing point of the material and the atmospheric pressure. For example, ; in They are all obtained by regression of experimental data, and are used to describe the positive or negative offset of the phase change temperature when the atmospheric pressure changes by one unit. In order to more accurately reflect the temperature drift of the material in the critical phase change range, piecewise functions can also be used in high altitude or low altitude sections to correct the special phase change behavior of the material in extreme environments. By combining the above model with the reference phase change temperature, a phase change point offset calculation model can be defined to dynamically calculate each critical air pressure value in subsequent steps, so as to obtain the actual phase change temperature offset of the material under different pressure conditions. For each air pressure value in the critical air pressure point sequence, the phase change point offset calculation model is used to calculate and obtain the discrete phase change point offset value. It is necessary to take out the critical air pressure points one by one in the program loop and substitute them as input variables into the above equation or fitting function. If the critical air pressure point sequence is set to , then for each When calculating, first determine the benchmark phase change temperature at that time or location. , and then introduce the corresponding parameter combination to obtain the offset of the phase change temperature by solving the equation If a difference between the measured pressure and the predicted value is observed in some sections, the external temperature, humidity and even wind speed factors can be added to the model as correction terms during the calculation. For example, a differential form or interpolation spline can be added to the main Clausius-Clapeyron equation to appropriately calibrate the offset in the real environment. After this treatment, for the same material base temperature and the same pressure level, if the additional climatic conditions are significantly different, the calculation results will be slightly corrected to be closer to the actual situation. The offset obtained in each calculation is recorded as , and the corresponding key pressure points To avoid abnormal results at high altitudes or other extreme conditions, the discrete values ​​need to be checked for validity, such as ensuring It will not be greater than a reasonable upper limit or lower than a reasonable lower limit. If some jump mutations appear in the discrete values, the sampling interval of the air pressure data can be checked or the thermodynamic characteristics of the phase change material in the Linyi pressure section can be checked to eliminate unreasonable values ​​caused by measurement errors or extreme temporary conditions. The generated discrete phase change point offset value sequence can be temporarily stored in the data structure and corresponded to the air pressure value with an index number, providing a numerical basis for the next step of establishing a continuous curve. The discrete phase change point offset values ​​are interpolated to obtain a continuous phase change point offset function, and the phase change point offset curve covering the entire air pressure range is generated based on the continuous phase change point offset function. The discrete data generated in the previous step will be The key control point is considered, and the interpolation algorithm is selected according to the requirements for temperature change smoothness and numerical accuracy. If simple linear interpolation cannot well capture the nonlinear drift of the material in certain pressure sections, a higher-order spline interpolation or piecewise polynomial fitting can be introduced to make ; in It is usually set to a cubic spline or a higher-order smooth function, and continuity and continuous differentiability are guaranteed in adjacent intervals. After completing the spline fitting, you can Any point on Query the phase change offset value , thus forming a three-dimensional phase transition shift function. Lower or lower than For higher extreme pressures, the interpolation function can be estimated by extrapolation or extension, but extrapolation errors should be treated with caution. A complete and smooth phase change point offset curve is obtained. The curve is marked according to time, location or other relevant dimensions, so that the actual phase change temperature of the material at different altitudes can be quickly inferred by comparing the air pressure distribution at each time or station in the transportation route. If it is necessary to query the phase change of the material in real time at the execution level, this offset curve can be embedded in the control system in the form of a data table or function, so that the coordinated scheduling of the compressor and phase change material is more targeted. By locating the corresponding air pressure value on the curve and retrieving the corresponding offset temperature, a complete set of dynamic temperature control strategies under different air pressures is formed, and refined phase change management capabilities are provided in plateaus, low pressure or other special environments.

[0042] 104. Carry out multimodal dynamic management of cold chain logistics process based on temperature control scheme and refrigeration strategy.

[0043] In one embodiment of the present invention, the multimodal dynamic management of the logistics process of the cold chain logistics according to the temperature control scheme and the refrigeration strategy includes: according to the temperature control scheme and the refrigeration strategy, real-time monitoring and analysis of the multimodal environmental parameters in the cold chain logistics process to obtain the current cold chain status evaluation result; based on the cold chain status evaluation result, combined with preset management rules, generating control instructions for the refrigeration system and phase change materials to obtain a resource allocation plan; converting the resource allocation plan into specific execution instructions, and adjusting the compressor speed, expansion valve opening, refrigerant ratio and activation of the phase change material of the refrigeration system in real time to obtain a dynamic control instruction sequence; according to the dynamic control instruction sequence, continuously monitoring and adjusting the environmental parameters in the cold chain logistics process, and recording the control effect to form a closed-loop feedback control to achieve multimodal dynamic management of cold chain logistics.

[0044] Specifically, according to the temperature control scheme and refrigeration strategy, the multimodal environmental parameters in the cold chain logistics process are monitored and analyzed in real time. When the current cold chain status evaluation results are obtained, it is necessary to deploy multi-source sensors that can sense parameters such as temperature, humidity, air pressure and light intensity at different stages of cold chain transportation, and integrate the collected original signals through a unified data acquisition module. In order to cope with the suddenness and diversity of environmental changes, it is necessary to rely on the multivariate time series prediction model and integrated learning correction mechanism established in the early stage, compare the sensor data with the existing prediction curve in the background system, check the difference between the current measured value and the predicted value, and identify whether abnormal working conditions such as high temperature, high humidity or sudden drop in air pressure occur. At the same time, it is also necessary to monitor the operating status of the actuators such as the phase change material module, compressor and expansion valve, such as whether the phase change material is in a molten state and whether the compressor speed is in a reasonable range. After inputting this information into the state analysis program, key indicators reflecting the overall health and temperature control level of the cold chain system can be extracted from it through feature extraction and pattern recognition technology, such as the deviation of the average temperature in the cabin from the target value and the length of time the humidity stays within the limit range. If the status analysis program finds that certain indicators are approaching or exceeding the threshold, it will mark the potential risks and attach related information, such as indicating which factors cause fluctuations. Then, the current cold chain status evaluation results with temperature stability, product safety and energy consumption rationality as the core are given by integrating environmental predictions, actual measurements and equipment operation data, and output in the form of data structures or visual monitoring interfaces. This process emphasizes the real-time tracking and traceability analysis capabilities of multimodal environmental parameters, enabling the system to perceive the differences between the outside and the inside at the first time, and match or correct the overall control strategy of cold chain transportation based on the evaluation results.

[0045] Specifically, based on the cold chain status evaluation results, combined with the preset management rules, control instructions for the refrigeration system and phase change materials are generated. When the resource allocation plan is obtained, it is necessary to extract the most urgent problems at the current stage from the evaluation results, such as the cabin temperature is close to the critical upper limit, the phase change material has not yet entered the expected phase change range, or the refrigerant ratio deviates from the optimal range. Management rules are often composed of multi-dimensional thresholds or expert system logic, and specific judgment conditions and scheduling principles can be written into a set of rule engines. If the evaluation results show that the cabin temperature is higher than the set value for several consecutive minutes, then under the retrieval of management rules, the processing strategy related to "high temperature overload" will be found first to determine how to increase the compressor speed in this situation or when to quickly trigger the release of more phase change materials to make up for the gap. If the cabin temperature fluctuates slightly but the air pressure changes dramatically, the phase change point offset of the phase change material will become an important consideration, and the management rules will determine whether it is necessary to adjust the timing or proportion of material activation to cope with the high altitude environment. In addition to the temperature dimension, management guidelines for humidity, energy consumption, and equipment load must also be considered. For example, if the humidity is too high in the evaluation report, a dehumidification module or ventilation device can be called to reduce the impact of water vapor on the quality of the goods. Ultimately, the decision content that meets these rule logics will be summarized into a list of control instructions, including the specific allocation amounts and execution order of the compressor, expansion valve, refrigerant ratio, and phase change material modules, and the corresponding priority will be given so that instructions can be issued quickly in a short time. At this time, a resource allocation plan will be generated to allocate the limited refrigeration capacity, phase change material heat storage and heat release potential to the required areas or time periods according to the optimal strategy, helping the goods to remain in the appropriate storage temperature range and taking into account energy consumption and efficiency goals.

[0046] Specifically, the resource allocation scheme is converted into specific execution instructions, and the compressor speed, expansion valve opening, refrigerant ratio and activation of the phase change material of the refrigeration system are adjusted in real time. When a dynamic control instruction sequence is obtained, a set of software and hardware linkage execution mechanisms needs to be established at the control level. The control system usually has a programmable logic controller or an embedded microcontroller unit to update the control command within a time step of seconds or shorter. The resource allocation scheme defines the ideal state of each stage, such as how the compressor speed should change with time or temperature curve, the range in which the expansion valve opening should be maintained, and when the refrigerant circulation line switches to a mixed refrigerant with a different ratio. If the phase change material needs to be activated to release the accumulated cold in a certain period of time, the control system will change the isolation valve or agitator of the material through the drive device, so that the phase change material is in contact with the outside air or the refrigerant to achieve rapid heat release. If the material is to be kept from being triggered in another period of time, the valve can be instructed to be in a half-closed or fully closed state to retain the phase change potential. The generation process of the execution instruction sequence takes into account the physical delay between each command being issued and its actual effectiveness, as well as the inertia of equipment such as compressors or phase change materials, to avoid issuing a large number of commands with large changes at the same time, which may cause system oscillation or equipment overload. Some systems also provide PID or more advanced adaptive control models for secondary correction of the instruction sequence, so that it can maintain fine-grained control of temperature, humidity, and air pressure during actual operation. When the instruction takes effect, the system will modify or continue to execute the original instruction sequence at any time according to the sensor data or status evaluation results of the next cycle, so that multimodal management has continuity and flexibility.

[0047] Specifically, according to the dynamic control instruction sequence, the environmental parameters in the cold chain logistics process are continuously monitored and adjusted, and the control effect is recorded to form a closed-loop feedback control. When realizing the multimodal dynamic management of cold chain logistics, it is necessary to let the monitoring module continuously obtain the latest multimodal data and compare it with the actual effect produced by the current execution instruction. If the temperature, humidity or air pressure in the cabin has changed significantly in the expected direction in a short period of time, this round of regulation can gradually converge; if the data shows that the effect is not good, or there are unpredictable and drastic fluctuations in the outside world, the closed-loop feedback control will immediately judge the instruction deviation and make corrections. The correction includes timely increase or decrease of the compressor speed, fine-tuning of the expansion valve opening, and re-evaluation of the refrigerant ratio and the activation state of the phase change material to ensure that the temperature curve in the cabin is steadily approaching the target value and minimizes overshoot or lag. In this process, the execution result of each regulation will be written into the system log together with the data read by the sensor. If it is monitored that the energy consumption or phase change material efficiency over a long period of time is not up to expectations, potential problems can be found through offline analysis and online intelligent algorithms, such as temporary changes in transportation routes or reduced equipment efficiency, and improved in the next cycle of the control loop. Through this real-time-historical fusion approach, multimodal dynamic management can continuously update and evolve itself, reduce energy waste while ensuring the quality and safety of goods, and bring higher operational reliability and visual management levels to cold chain logistics.

[0048] In this embodiment, by dynamically monitoring and predicting the multimodal environmental parameters inside and outside the cold chain container, the prediction results of environmental parameter changes within a preset time range are obtained; according to the prediction results of environmental parameter changes and the logistics route, the compressor speed, expansion valve opening and refrigerant ratio of the refrigeration system are dynamically adjusted to obtain the refrigeration strategy of the refrigeration system; based on the refrigeration strategy and logistics route, the activation timing and proportion of the phase change material are calculated and controlled to obtain a temperature control scheme that takes into account the phase change point offset under different altitude pressures; according to the temperature control scheme and refrigeration strategy, the cold chain logistics process is managed in a multimodal dynamic manner. The present invention can effectively cope with complex environmental changes in the multimodal transport process, dynamically adjust the refrigeration strategy and the use of phase change materials, and improve the quality and reliability of cold chain logistics.

[0049] The above describes the multimodal management method for cold chain logistics based on artificial intelligence in an embodiment of the present invention. The following describes the multimodal management system for cold chain logistics based on artificial intelligence in an embodiment of the present invention. Figure 2 In one embodiment of the present invention, an embodiment of a multimodal management system for cold chain logistics based on artificial intelligence includes: The environmental prediction module 201 is used to dynamically monitor and predict the multimodal environmental parameters inside and outside the cold chain container during the cold chain logistics process, and obtain the prediction results of environmental parameter changes within a preset time range; A refrigeration control module 202 is used to dynamically control the compressor speed, expansion valve opening and refrigerant ratio of the refrigeration system according to the prediction result of the environmental parameter change and the logistics route of the cold chain logistics, so as to obtain a refrigeration strategy for the refrigeration system; The phase change management module 203 is used to calculate and control the activation timing and proportion of the phase change material in the cold chain container based on the refrigeration strategy and the logistics route, and obtain a temperature control solution that takes into account the phase change point offset under different altitude pressures; The dynamic management module 204 is used to perform multi-modal dynamic management of the cold chain logistics process according to the temperature control scheme and the refrigeration strategy.

[0050] In an embodiment of the present invention, the artificial intelligence-based cold chain logistics multimodal management system runs the artificial intelligence-based cold chain logistics multimodal management method described above, and the artificial intelligence-based cold chain logistics multimodal management system dynamically monitors and predicts the multimodal environmental parameters inside and outside the cold chain container to obtain the prediction results of environmental parameter changes within a preset time range; according to the prediction results of environmental parameter changes and the logistics route, the compressor speed, expansion valve opening and refrigerant ratio of the refrigeration system are dynamically regulated to obtain the refrigeration strategy of the refrigeration system; based on the refrigeration strategy and logistics route, the activation timing and proportion of the phase change material are calculated and controlled to obtain a temperature control scheme that considers the phase change point offset under different altitude pressures; according to the temperature control scheme and refrigeration strategy, the cold chain logistics process is dynamically managed in a multimodal manner. The present invention can effectively cope with complex environmental changes in the multimodal transport process, dynamically adjust the refrigeration strategy and the use of phase change materials, and improve the quality and reliability of cold chain logistics.

[0051] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the above-described system, device, or unit can refer to the corresponding process in the aforementioned method embodiment and will not be repeated here.

[0052] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or the whole or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program code.

[0053] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A multimodal management method for cold chain logistics based on artificial intelligence, characterized in that: The multimodal management method for cold chain logistics based on artificial intelligence includes: Dynamically monitor and predict the multimodal environmental parameters inside and outside the cold chain container during the cold chain logistics process to obtain the prediction results of environmental parameter changes within a preset time range; According to the prediction results of the environmental parameter changes and the logistics route of the cold chain logistics, the compressor speed, expansion valve opening and refrigerant ratio of the refrigeration system are dynamically controlled to obtain a refrigeration strategy for the refrigeration system; Based on the refrigeration strategy and the logistics route, the activation timing and proportion of the phase change material in the cold chain container are calculated and controlled to obtain a temperature control solution that takes into account the phase change point offset under different altitude pressures; The cold chain logistics process is managed dynamically in a multi-modal manner according to the temperature control scheme and the refrigeration strategy.

2. The multimodal management method for cold chain logistics based on artificial intelligence according to claim 1 is characterized in that: The dynamic monitoring and prediction of multimodal environmental parameters inside and outside the cold chain container during the cold chain logistics process to obtain the prediction results of environmental parameter changes within a preset time range include: Perform multi-frequency sampling processing on the temperature, humidity, air pressure and light intensity inside and outside the cold chain container during the cold chain logistics process to obtain the original environmental parameter data set; Performing outlier detection and data cleaning processing on the original environmental parameter data set to obtain a preprocessed environmental parameter data set; Inputting the preprocessed environmental parameter data set into a preset multivariate time series prediction model to obtain preliminary environmental parameter prediction results; Using a preset integrated learning model and historical environmental data to perform prediction correction on the preliminary environmental parameter prediction results to obtain a corrected environmental parameter prediction result; The corrected environmental parameter prediction results are subjected to time series interpolation processing to obtain environmental parameter change prediction results within a preset time range.

3. The multimodal management method for cold chain logistics based on artificial intelligence according to claim 1 is characterized in that: According to the prediction result of the environmental parameter change and the logistics route of the cold chain logistics, the compressor speed, expansion valve opening and refrigerant ratio of the refrigeration system are dynamically controlled to obtain the refrigeration strategy of the refrigeration system, including: According to the prediction results of the environmental parameter changes, the expected temperature difference and the required refrigeration capacity in each transportation stage are calculated to obtain a refrigeration demand curve; Segmenting the logistics route, identifying key environmental change points, and obtaining a route feature sequence; Calculating the minimum power requirement of the compressor of the refrigeration system according to the refrigeration demand curve and the route characteristic sequence to obtain a basic speed curve of the compressor; Based on the compressor basic speed curve and the environmental parameter change prediction result, the compressor speed is dynamically adjusted to obtain a compressor speed control sequence; Calculating the optimal opening of the expansion valve according to the compressor speed control sequence and the required refrigeration capacity to obtain the expansion valve opening control sequence; Based on the prediction results of the environmental parameter changes, the ideal refrigerant state under different environments is calculated to obtain a refrigerant ratio adjustment plan; The compressor speed control sequence, the expansion valve opening control sequence and the refrigerant ratio adjustment scheme are time-synchronized to obtain a refrigeration strategy for the refrigeration system.

4. The multimodal management method for cold chain logistics based on artificial intelligence according to claim 3 is characterized in that: The step of calculating the minimum power requirement of the compressor of the refrigeration system according to the refrigeration demand curve and the route characteristic sequence to obtain the basic speed curve of the compressor includes: The refrigeration demand curve is subjected to piecewise linearization processing to obtain a piecewise refrigeration demand function, and the environmental characteristics of each transportation stage are quantitatively scored according to the route characteristic sequence to obtain an environmental complexity index sequence; The segmented refrigeration demand function is weightedly superimposed with the environmental complexity index sequence to obtain a comprehensive refrigeration load curve; According to the comprehensive refrigeration load curve, the theoretical minimum compression work is calculated using a thermodynamic formula to obtain a compression power curve, and the compression power curve is converted into a compressor speed to obtain a compressor basic speed curve.

5. The multimodal management method for cold chain logistics based on artificial intelligence according to claim 1 is characterized in that: The activation timing and proportion of the phase change material in the cold chain container are calculated and controlled based on the refrigeration strategy and the logistics route to obtain a temperature control solution that takes into account the phase change point offset under different altitude pressures, including: Extracting the altitude change sequence according to the logistics route, calculating the air pressure change curve during the cold chain logistics process, and obtaining the air pressure distribution in different transportation stages; Based on the air pressure distribution and the reference phase change temperature of the phase change material in the cold chain container, the phase change point offset under different air pressures is calculated to obtain a phase change point offset curve; Compare and analyze the refrigeration strategy with the phase change point offset curve to determine the activation timing and usage ratio of the phase change material and obtain a preliminary phase change material usage plan; According to the preliminary phase change material usage plan and refrigeration strategy, performance matching and optimization adjustments are carried out to obtain a temperature control plan that takes into account the phase change point offset under different altitude pressures.

6. The multimodal management method for cold chain logistics based on artificial intelligence according to claim 5 is characterized in that: The calculating the phase change point offset under different air pressures based on the air pressure distribution and the reference phase change temperature of the phase change material in the cold chain container to obtain the phase change point offset curve comprises: Discretizing the air pressure distribution to obtain a key air pressure point sequence, and establishing an air pressure-temperature relationship equation based on a reference phase change temperature of the phase change material to obtain a phase change point offset calculation model; For each air pressure value in the key air pressure point sequence, the phase change point offset calculation model is used to calculate to obtain a discrete phase change point offset value; The discrete phase change point offset values ​​are interpolated to obtain a continuous phase change point offset function, and a phase change point offset curve covering the entire air pressure range is generated according to the continuous phase change point offset function.

7. The multimodal management method for cold chain logistics based on artificial intelligence according to claim 1 is characterized in that: The multi-modal dynamic management of the logistics process of the cold chain logistics according to the temperature control scheme and the refrigeration strategy includes: According to the temperature control scheme and refrigeration strategy, the multimodal environmental parameters in the cold chain logistics process are monitored and analyzed in real time to obtain the current cold chain status evaluation results; Based on the cold chain status assessment result and in combination with preset management rules, control instructions for the refrigeration system and phase change materials are generated to obtain a resource allocation plan; The resource allocation scheme is converted into specific execution instructions, and the compressor speed, expansion valve opening, refrigerant ratio and activation of phase change materials of the refrigeration system are adjusted in real time to obtain a dynamic control instruction sequence; According to the dynamic control instruction sequence, the environmental parameters in the cold chain logistics process are continuously monitored and adjusted, and the regulation effect is recorded to form a closed-loop feedback control to achieve multi-modal dynamic management of cold chain logistics.

8. A multimodal management system for cold chain logistics based on artificial intelligence, characterized in that: The artificial intelligence-based cold chain logistics multimodal management system includes: The environmental prediction module is used to dynamically monitor and predict the multimodal environmental parameters inside and outside the cold chain container during the cold chain logistics process, and obtain the prediction results of environmental parameter changes within a preset time range; A refrigeration control module is used to dynamically control the compressor speed, expansion valve opening and refrigerant ratio of the refrigeration system according to the prediction results of the environmental parameter changes and the logistics route of the cold chain logistics, so as to obtain a refrigeration strategy for the refrigeration system; A phase change management module, for calculating and controlling the activation timing and proportion of the phase change material in the cold chain container based on the refrigeration strategy and the logistics route, and obtaining a temperature control solution that takes into account the phase change point offset under different altitude pressures; A dynamic management module is used to perform multimodal dynamic management of the cold chain logistics process according to the temperature control scheme and the refrigeration strategy.

Images