Intelligent refrigerator monitoring method based on unmanned aerial vehicle transportation and refrigerator

Abstract

The invention relates to the field of cold-chain logistics transportation, and discloses an intelligent refrigerator monitoring method based on unmanned aerial vehicle transportation and a refrigerator, and the method comprises the steps: collecting the flight state and environment data, and calculating the surface radiant heat flux of the refrigerator through a geometric projection model; extracting high-frequency disturbance characteristics of rotor motor current as flow field turbulivity characterization quantities, and performing in-situ correction on the reference convective heat transfer coefficient to calculate convective heat flux; predicting the temperature rise rate in the refrigerator based on the synthesized total external heat flux; and when the temperature rise exceeds the standard, constructing an anisotropic thermal impedance potential field containing thermal cost and flight energy consumption cost, solving an optimal attitude by using a gradient optimization algorithm, and driving the unmanned aerial vehicle to avoid a high-heat area. The invention further provides a refrigerator which adopts the Stirling refrigeration compressor and the contact type refrigeration end. According to the method, the problems of thermal load evaluation distortion caused by neglecting of rotor turbulence and high cargo overtemperature risk caused by lack of an active attitude avoidance mechanism are solved, and collaborative optimization of thermal protection and flight energy efficiency is realized.

Description

Technical Field

[0001] This invention relates to the field of cold chain logistics transportation technology, specifically to an intelligent monitoring method for refrigerated containers based on drone transportation and the refrigerated container itself. Background Technology

[0002] With the development of low-altitude logistics networks, the use of drones for cold chain transportation of temperature-sensitive goods such as blood products, biological samples, and high-end fresh produce has entered the application stage. Current drone cold chain transportation systems mainly consist of a drone flight platform and an onboard refrigerated container. The refrigerated container typically uses high thermal resistance vacuum insulation panels or polyurethane foam as insulation layers and is equipped with temperature sensors and small refrigeration modules. During transportation missions, existing monitoring methods primarily rely on collecting real-time temperature data inside the refrigerated container to determine the cargo's condition. When a temperature rise is detected, the passive insulation properties of the insulation material are used to slow heat transfer, or the onboard refrigeration equipment is activated for active cooling to maintain the container's temperature within a preset safe range. The drone flies along a predetermined route, its flight attitude and speed primarily determined by navigation based on path planning, while the refrigerated container, as a payload, moves along with the drone.

[0003] However, the existing technologies have the following limitations in practical applications: First, existing heat load assessment methods mainly rely on static or advection-based aerodynamic formulas to calculate the convective heat transfer coefficient, ignoring the strong high-frequency turbulence characteristics of the downwash flow field generated by the high-speed rotation of the UAV rotor. This turbulence can damage the thermal boundary layer on the surface of the freezer, resulting in the actual convective heat flux being much higher than the theoretical calculation value. Consequently, the system cannot accurately sense and predict the temperature rise trend inside the freezer. Second, existing protection measures rely entirely on insulation materials or limited cooling power for passive defense, lacking a mechanism to actively avoid environmental thermal risks by utilizing the UAV's own maneuverability. When encountering strong solar radiation at a specific angle or high-temperature airflow, the UAV cannot adjust its attitude to reduce the heated area of ​​the freezer, leading to the risk of overheating damage to the goods. Finally, existing monitoring logic does not consider the coupling relationship between thermal protection and flight energy consumption, lacking overall optimization of the flight attitude energy efficiency ratio. This can easily lead to the UAV adopting an inappropriate flight attitude, causing the battery to be consumed too quickly, thus affecting the transportation range or causing power safety issues. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides an intelligent monitoring method and a refrigerator based on drone transportation. This method solves the problems in existing drone cold chain transportation, which are caused by the inability to accurately quantify the thermal effects of rotor turbulence flow field and the lack of active mechanisms to avoid high heat flux, resulting in large deviations in the predicted temperature rise inside the refrigerator and a high risk of goods overheating.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] The first aspect of this invention provides a method for intelligent monitoring of refrigerated containers based on unmanned aerial vehicle (UAV) transportation, the method comprising the following steps:

[0007] First, real-time data is collected on the drone's flight status, environmental parameters, and the freezer's geometric and physical parameters. The flight status data includes Euler angles, three-dimensional position, ground velocity vector, and rotor motor operating parameters; the environmental parameter data includes ambient temperature and relative airflow vector.

[0008] Secondly, radiative heat flux is calculated based on the dynamic attitude changes during flight. A geometric projection model is constructed using the UAV's Euler angles and the current solar incidence vector. This model is then used to calculate the effective light-receiving area of ​​each surface of the freezer under the current attitude. Based on this, and combined with the solar radiation intensity, the radiative heat flux acting on the freezer surface is calculated. This step corrects for the changes in light-receiving area caused by the UAV's attitude tilt through geometric projection, achieving dynamic and accurate quantification of the radiative heat load.

[0009] Furthermore, convective heat flux was calculated to address the nonlinear influence of the rotor downwash flow field on convective heat transfer. High-frequency disturbance characteristics of the instantaneous current values ​​of each UAV rotor motor were extracted and used as an observation variable characterizing the turbulence intensity of the rotor downwash flow field. This observation variable was used to perform in-situ correction on the baseline convective heat transfer coefficient calculated based on relative air velocity, thereby obtaining the actual convective heat transfer coefficient of the cooler surface that conforms to actual flight conditions, and the convective heat flux was calculated accordingly. This approach establishes a mapping relationship between motor current fluctuations and the intensity of turbulence in the flow field, solving the problem that traditional laminar or steady-state turbulence models cannot accurately describe the heat transfer characteristics of the complex rotor flow field.

[0010] Subsequently, the total external heat flux is synthesized and temperature changes are predicted. Radiative and convective heat fluxes are vector-synthesized, and combined with the thermal resistance parameters of the freezer's insulation layer and the convective thermal resistance of its inner surface, a heat transfer model is constructed using the lumped parameter method to predict the rate of temperature change inside the freezer within a preset future time period.

[0011] Finally, active thermal avoidance control based on potential field optimization is implemented. When the predicted rate of temperature change exceeds a preset threshold, an anisotropic thermal impedance flight potential field is constructed. This potential field integrates thermal costs (external heat flux) and flight energy costs (aerodynamic power and hovering power). A gradient optimization algorithm is used within this potential field to find the optimal target attitude that minimizes the total cost, and control commands are generated based on this attitude to drive the UAV to adjust its yaw or tilt angle. This mechanism, while ensuring mission continuity, achieves autonomous avoidance of high heat flux regions by changing the fuselage's attitude relative to the sun and wind fields.

[0012] The specific calculation logic for radiative heat flux is as follows: The solar unit direction vector in the geodetic coordinate system is transformed to the machine coordinate system. The dot product of the solar incident vector and the outer surface normal vector is calculated across all surfaces of the freezer. The sun-facing surface with a positive dot product is selected, and the dot product value, surface physical area, solar radiation absorptivity, and solar irradiance are multiplied to obtain the radiative heat flux per surface. Finally, the total radiative heat flux is obtained by summing the values ​​of all sun-facing surfaces.

[0013] The logic for extracting high-frequency current disturbance characteristics and correcting the convective heat transfer coefficient is as follows: A sampling window is set to acquire the instantaneous phase current sequence. The square of the difference between each data point and the average current value is calculated. Statistical processing yields the root mean square value, reflecting the severity of current fluctuations, i.e., the current disturbance index. During the correction process, a baseline convective heat transfer coefficient is calculated based on the hydrodynamic flat-plate flow theory, and a correction factor incorporating a flow-thermal coupling weighting coefficient is constructed. This correction factor is related to the ratio of the current disturbance index to the total average current. By multiplying the baseline coefficient by the correction factor, dynamic compensation of the convective heat transfer coefficient is achieved.

[0014] The prediction logic for the temperature change rate is as follows: Calculate the total thermal resistance from the external environment to the interior of the freezer. This resistance is composed of the external convective thermal resistance, the insulation thermal resistance, and the internal convective thermal resistance connected in series. Calculate the transmission component of the total external heat flux and the heat transfer component due to the ambient temperature difference. Subtract the rated cooling power of the internal cold source from the sum of these two components to obtain the net heat input. Divide the net heat input by the combined specific heat capacity of the freezer and its load to calculate the temperature change rate.

[0015] The logic for constructing and optimizing the flight potential field is as follows: A thermal impedance function and a flight energy consumption impedance function are established. The thermal impedance function is associated with the thermal cost weight and the predicted heat flux; the flight energy consumption impedance function is associated with the energy consumption weight and the estimated total flight power, where the aerodynamic power term is proportional to the square of the relative air velocity and the aerodynamic drag coefficient that varies with attitude. The sum of these two functions constitutes the total potential field cost function. By calculating the gradient vector of this cost function with respect to the attitude angle variable, and iteratively updating the attitude vector along the negative gradient direction until the gradient magnitude converges, the optimal target attitude that balances thermal protection and flight energy consumption is determined.

[0016] A second aspect of the present invention provides a freezer that is applied to the above-described intelligent monitoring method.

[0017] The main structure of the freezer includes a cabinet, which is constructed as a hollow, enclosed enclosure with walls filled with an insulation layer to insulate against the external heat. A refrigeration compressor is fixedly installed inside the freezer to provide an active cooling source.

[0018] The refrigeration compressor is equipped with a refrigeration end extending along one side, and a refrigeration terminal is connected to the end of the refrigeration end. The refrigeration terminal maintains physical contact with the heat-conducting wall surface of the cabinet, establishing a cold energy transfer channel from the refrigeration compressor to the inside of the cabinet through contact heat conduction.

[0019] In addition, the freezer is equipped with temperature sensors inside the insulation layer or within the storage space to collect the internal temperature of the freezer in real time and generate data feedback.

[0020] Furthermore, the refrigeration compressor is preferably a Stirling refrigeration compressor. This type of compressor has the ability to maintain continuous and stable refrigeration operation even under conditions of drastic changes in the flight attitude of the UAV or multi-angle shaking. At the same time, the refrigeration end is made of a high thermal conductivity metal material, and its geometry is adapted to the contact area with the cabinet to maximize the contact heat transfer area, ensuring efficient conduction of cold air.

[0021] This invention provides a method for intelligent monitoring of refrigerated containers based on drone transportation, and a refrigerated container itself. It has the following beneficial effects:

[0022] 1. This invention characterizes the turbulence intensity of the downwash flow field by extracting the high-frequency disturbance characteristics of the rotor motor's current, and uses this to perform in-situ correction of the reference convective heat transfer coefficient on the surface of the freezer. This approach establishes a mapping relationship between the characteristics of the UAV's power system and the thermal boundary conditions of the freezer, solving the problem that existing monitoring methods rely solely on static aerodynamic formulas and cannot detect the destruction and reconstruction of the thermal boundary layer on the freezer surface caused by severe rotor disturbances, thus leading to large deviations in the predicted temperature rise inside the freezer. This improves the confidence level of the freezer thermal monitoring model.

[0023] 2. This invention constructs a thermal impedance potential field and monitors the rate of internal temperature change. When a risk of excessive temperature rise is detected, it utilizes the asymmetry of the refrigerator's geometry to adjust the transport posture by outputting commands, thereby changing the effective light-receiving surface and windward surface of the refrigerator relative to the sun and airflow. This allows for parallel monitoring and control, solving the problem that traditional cold chain transportation relies entirely on insulation materials for heat insulation and cannot cope with sudden high radiation or strong convection environments that cause goods to overheat and be damaged. This improves the transportation safety of high-value cold chain goods.

[0024] 3. This invention incorporates flight energy consumption costs into the potential field function and weights them together with thermal costs during the generation of thermal avoidance strategies. This monitoring logic ensures that when planning the attitude of the refrigerated container, the transport vehicle will not be forced into a high-drag or high-power state in pursuit of extremely low heat input. This resolves the technical contradiction that a single-dimensional thermal monitoring logic might lead to premature battery depletion, thus preventing the completion of the entire transport mission, and guarantees the engineering practicality of the refrigerated container monitoring method. Attached Figure Description

[0025] Figure 1 This is a flowchart of the method steps of the present invention; Figure 2 This is a structural distribution diagram of the freezer of the present invention; The components include: 1. Cabinet; 2. Refrigeration compressor; 3. Refrigeration end; 4. Refrigeration terminal; 5. Insulation layer; and 6. Temperature sensor. Detailed Implementation

[0026] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0027] Example 1:

[0028] Please see the appendix Figure 1 This invention provides a method for intelligent monitoring of refrigerated containers based on drone transportation, comprising the following steps:

[0029] S1. The onboard computing equipment collects multi-source heterogeneous data in real time through a data interface. This multi-source heterogeneous data includes the UAV's flight status data, environmental parameter data, and the freezer's geometric and physical parameter data. Flight status data includes at least the UAV's Euler angles, three-dimensional position coordinates, ground speed vector, real-time rotational speed of each rotor motor, and instantaneous current value of each rotor motor at the current moment. Environmental parameter data includes at least the ambient temperature and relative wind speed vector.

[0030] S2. The onboard computing equipment constructs a geometric projection model based on the UAV's Euler angles and the current solar incidence vector. Using this model, it calculates the effective light-receiving area of ​​each surface of the freezer in its current orientation and, combined with the solar radiation intensity, calculates the radiative heat flux acting on the freezer surface.

[0031] S3. The airborne computing equipment extracts high-frequency current disturbance characteristics based on the instantaneous current values ​​of each rotor motor. These high-frequency current disturbance characteristics are used as observation variables characterizing the turbulence intensity of the rotor downwash flow field. The baseline convective heat transfer coefficient calculated based on relative wind speed is then corrected in situ to obtain the actual convective heat transfer coefficient of the freezer surface, and the convective heat flux of the freezer surface is calculated accordingly.

[0032] S4. The onboard computing equipment performs vector synthesis of radiative and convective heat flux to obtain the total external heat flux acting on the freezer. Based on the total external heat flux and the thermal resistance parameters of the freezer's insulation layer, the rate of change of the freezer's internal temperature within a preset future time period is predicted using heat transfer differential equations.

[0033] S5. When the predicted rate of temperature change exceeds a preset threshold, the airborne computing equipment constructs an anisotropic thermal impedance flight potential field that includes thermal costs and flight energy consumption costs. In this anisotropic thermal impedance flight potential field, a gradient optimization algorithm is used to solve for the optimal target attitude that minimizes the total cost. Based on the optimal target attitude, flight control commands are generated and sent to the flight controller to drive the UAV to adjust its yaw or tilt angle to avoid high heat flux regions.

[0034] In the above method, steps S1-S5 constitute a closed-loop control process. The onboard computing equipment periodically executes the above steps throughout the entire flight of the UAV, adjusting the UAV's flight attitude in real time according to the dynamic changes in the external aerodynamic and thermal environment, thereby reducing the cooling loss of the refrigerator while maintaining the flight mission.

[0035] As the implementing entity of the method, the airborne computing device, during the monitoring process of the UAV refrigerated container transportation mission, first establishes a reference coordinate system to describe the spatial position and physical quantity direction, and digitally models the geometric and physical parameters of the refrigerated container. Then, it reads the flight status and environmental parameters in real time through a data bus. The specific implementation process includes the following sub-steps:

[0036] S11. Establish a reference coordinate system and attitude description model.

[0037] To accurately calculate the vector form of heat flux and aerodynamic flow field, the airborne computing device defines two Cartesian coordinate systems: a geodetic coordinate system and a body coordinate system. The geodetic coordinate system has its origin at the UAV's takeoff point and includes the XE axis pointing due north, the YE axis pointing due east, and the ZE axis perpendicular to the Earth's center. The body coordinate system has its origin at the UAV's center of mass and includes the XB axis pointing directly forward of the nose, the YB axis pointing to the right side of the fuselage, and the ZB axis perpendicular to the fuselage downwards.

[0038] The attitude of the drone in the air is described by an Euler angle vector, which contains three components:

[0039] ;

[0040] In the formula, The roll angle represents the angle of rotation of the fuselage around the XB axis. The pitch angle represents the rotation angle of the aircraft around the YB axis; Yaw angle represents the rotation angle of the airframe around the ZB axis.

[0041] The rotation transformation matrix for any vector from the geodetic coordinate system to the body coordinate system is determined by Euler angles, and its mathematical expression is denoted as:

[0042] ;

[0043] In the formula, , , These are the basic rotation matrices about each axis; This is the rotation transformation matrix. The specific numerical calculation methods for this matrix are standard techniques in rigid body dynamics, and their specific matrix expansions will not be listed here.

[0044] S12. Construct a geometric and thermal property parameter model for the freezer.

[0045] The onboard computing device pre-stores the geometric model data of the freezer. In this embodiment, the freezer is abstracted as a closed geometry with six surfaces. For the j-th surface of the freezer, its geometric features are described by its surface area and the normal vector of its outer surface. j takes the value of an integer from 1 to 6, corresponding to the front, back, left, right, top, and bottom surfaces of the freezer, respectively. Since the freezer is rigidly connected to the UAV frame, the normal vector of its outer surface is a constant vector in the body coordinate system. For example, if the front surface of the freezer faces the nose of the UAV, the normal vector of that surface is represented in the body coordinate system as follows: Meanwhile, the onboard computing equipment stores the thermal properties of the freezer's insulation layer, including the thermal conductivity of the insulation material, the thickness of the insulation layer, and the combined specific heat capacity of the freezer and the cargo inside.

[0046] S13. Real-time acquisition of flight status and dynamic response data.

[0047] The onboard computing equipment connects to the UAV's flight controller and electronic speed controller via a communication interface, acquiring flight status data in real time at a preset sampling frequency. The flight status data specifically includes:

[0048] The current Euler angles of the drone will be used for subsequent calculations of the solar incidence angle and flow field direction;

[0049] The three-dimensional position coordinates of the drone are used to determine its current latitude, longitude, and altitude to support the calculation of the sun's position;

[0050] The ground velocity vector of the drone is used for navigation calculations;

[0051] Operating status data for each rotor motor.

[0052] For a UAV with K rotors, the onboard computing device collects the rotational speed and instantaneous phase current of the k-th motor at time t. It should be noted that the instantaneous phase current collected in this embodiment is unsmoothed, high-frequency raw data, used in subsequent steps to extract characteristic quantities reflecting airflow disturbances.

[0053] S14. Obtain environmental parameter data.

[0054] Airborne computing devices acquire external environmental parameters through airborne sensors or data links. These environmental parameters include:

[0055] Ambient temperature was obtained by measuring the airborne temperature sensor.

[0056] The relative air velocity vector represents the air velocity relative to the UAV body. It can be directly measured by the airspeed tube on the air, or obtained by vector calculation using the ground velocity vector and wind speed observation data.

[0057] The current Coordinated Universal Time (UTC) is used to calculate the solar altitude angle and azimuth angle by combining latitude and longitude.

[0058] After completing the acquisition and initialization of multi-source data, the airborne computing equipment calculates the solar radiation heat load on each surface of the freezer based on the UAV's current flight attitude and geographical location using a geometric projection algorithm. This process is specifically implemented through the following sub-steps:

[0059] S21. Solve for the solar incident vector in the body coordinate system.

[0060] The onboard computing equipment first calculates the azimuth and altitude angles of the sun relative to the observer at the current moment, based on the UAV's current longitude, latitude, and Coordinated Universal Time (UTC), using a solar position algorithm. For the specific algorithm used to calculate the sun's position, those skilled in the art can employ astronomical algorithms (such as the SPA algorithm), which are well-known technologies and will not be elaborated upon here.

[0061] Based on the azimuth and elevation angles, the airborne computing equipment constructs a solar unit direction vector in the geodetic coordinate system. Then, using the rotation transformation matrix determined in step S11, the solar unit direction vector is projected onto the aircraft coordinate system to obtain the solar incident vector in the aircraft coordinate system. The expression for this vector is:

[0062] ;

[0063] In the formula, This is the solar incidence vector in the body coordinate system; To be based on the current Euler angles A defined rotation matrix from the geodetic coordinate system to the body coordinate system; This is the solar unit direction vector in the geodetic coordinate system.

[0064] S22. Calculate the effective radiative heat flux on one side.

[0065] After determining the direction of sunlight relative to the aircraft, the onboard computing equipment traverses every surface of the freezer, calculating the effective light-receiving area and radiative heat flux of each surface. For the j-th surface of the freezer, the intensity of solar radiation received depends on the geometric angle between the surface's normal vector and the solar incident vector. The onboard computing equipment reads the outer surface normal vector of the j-th surface of the freezer in the aircraft coordinate system and calculates its dot product with the solar incident vector. Since solar radiation cannot penetrate the freezer body to reach the shaded side, a non-negative constraint function is introduced during the calculation process, accumulating heat only on the sun-facing side. The instantaneous radiative heat flux of the j-th surface is calculated according to the following formula:

[0066] ;

[0067] In the formula, For the freezer The solar radiation heat flux absorbed by each surface; The solar radiation absorption rate of the outer surface of the freezer depends on the material and coating characteristics of the freezer surface. This represents the current solar irradiance, a value that can be directly measured by an airborne light sensor or estimated based on geographical location and weather models. For the freezer The physical area of ​​each surface; This represents the vector dot product of the solar incident vector and the surface normal vector; This is a function that maximizes the value of a component, used to remove the backlight component whose dot product is negative.

[0068] S23, Total radiant heat load of the synthetic refrigerator.

[0069] After calculating the radiation components of all external surfaces of the freezer, the onboard computing equipment sums the components as a scalar to obtain the total radiative heat load acting on the entire freezer at the current moment. The calculation formula is as follows:

[0070] ;

[0071] In the formula, This represents the total solar radiation heat flux received by the freezer. This represents the total number of outer surfaces of the freezer; for a standard rectangular freezer, this value is 6. For the freezer The solar radiation heat flux absorbed by a surface.

[0072] Through the above steps, the onboard computing equipment can quantify the differences in solar radiation intensity received by the freezer under different flight attitudes of the drone (such as side-flying and pitch-flying). For example, when the drone adjusts its yaw angle so that the smallest area of ​​the freezer (such as the side) faces the sun, and the largest area (such as the top or front) is in its own shadow or parallel to the light, the calculated... The value will decrease. This calculation result will serve as one of the basic input variables for the subsequent construction of the thermal impedance potential field.

[0073] During actual flight, the high-speed downwash airflow generated by the UAV rotor superimposed on the ambient natural wind field, forming a complex turbulent field with highly unsteady characteristics on the surface of the freezer. To accurately quantify the impact of this flow field on the convective heat transfer of the freezer surface, the onboard computing equipment does not simply rely on static aerodynamic formulas, but instead introduces the electrical response characteristics of the power system for in-situ correction in step S3. This process is specifically implemented through the following sub-steps:

[0074] S31. Construct a benchmark convection heat transfer model.

[0075] The airborne computing equipment first calculates the baseline convective heat transfer coefficient under ideal laminar flow conditions based on the flat plate grazing flow theory in fluid mechanics. Using the relative air velocity vector obtained in the preceding steps, the Reynolds number of the airflow across the freezer surface is calculated, and combined with the Prandtl number, the baseline convective heat transfer coefficient is derived using the Nusselt number correlation. Its calculation logic satisfies the following relationship:

[0076] ;

[0077] In the formula, The reference convective heat transfer coefficient; The thermal conductivity of air; The characteristic length of the freezer surface; is the Reynolds number, which represents the ratio of inertial force to viscous force; Prandtl number, which characterizes the ratio of a fluid's momentum diffusivity to its thermal diffusivity; , , As empirical constants related to the fluid flow state, the airborne computing equipment automatically selects the corresponding constant values ​​based on the current Reynolds number range.

[0078] S32. Extract the high-frequency disturbance characteristics of the motor current.

[0079] When a drone is in a turbulent flow field, the flight controller's attitude loop and angular velocity loop frequently adjust the motor output torque to maintain fuselage stability. This rapid change in torque directly translates into minute fluctuations in motor current. The onboard computing equipment captures the instantaneous phase current data stream of each rotor motor within a preset time window and calculates the high-frequency components of the current fluctuations. In this embodiment, the root mean square value of the current fluctuation is used as a quantitative indicator characterizing the intensity of turbulence in the flow field, defined as the current disturbance index. The calculation formula is as follows:

[0080] ;

[0081] In the formula, The current disturbance index; This represents the total number of rotor motors in the drone; This is the length of the sampling time window; For the first Each motor The instantaneous current value at a given moment; For the first The average current value of each motor within this time window. This index The larger the value, the more intense the impact and disturbance of the airflow on the rotor and fuselage, the higher the corresponding turbulence of the flow field, and the stronger the destructive effect of the airflow on the boundary layer of the freezer surface.

[0082] S33, In-situ correction of convective heat transfer coefficient.

[0083] The airborne computing equipment uses the extracted current disturbance index to compensate and correct the reference convective heat transfer coefficient, establishing a mapping relationship between the motor dynamic response and surface thermodynamic properties to obtain the equivalent convective heat transfer coefficient. The correction model is as follows:

[0084] ;

[0085] In the formula, The actual convective heat transfer coefficient of the freezer surface is corrected; The reference convective heat transfer coefficient is calculated in step S31; This is the heat transfer coupling weighting coefficient, which is a preset constant used to adjust the weight of the disturbance on the heat transfer coefficient. The current disturbance index is calculated in step S32; This is the total average current value of all motors within the current time window, used to normalize the disturbance.

[0086] Through this correction step, the airborne computing equipment can identify the difference in heat exchange effects between high wind speed and low turbulence (such as smooth cruise) and high wind speed and high turbulence (such as violent maneuvering or being disturbed by gusts), thus more accurately reflecting the actual heat dissipation or heat absorption conditions of the freezer surface.

[0087] S34. Calculate the total convective heat flux.

[0088] After determining the equivalent convective heat transfer coefficient that conforms to the current flow field characteristics, the onboard computing equipment calculates the convective heat transfer between each surface of the freezer and the ambient air according to Newton's law of cooling, and then sums them up to obtain the total convective heat flux:

[0089] ;

[0090] In the formula, This represents the total convective heat flux received by the freezer. This represents the total number of outer surfaces of the freezer; This is the corrected equivalent convective heat transfer coefficient; For the freezer The area of ​​each surface; The ambient temperature; For the freezer The outer wall temperature of each surface. It should be noted that when the ambient temperature... Temperature higher than the outer wall of the freezer When the airflow is at a certain temperature, it heats the freezer; conversely, it cools it. The onboard computing equipment automatically determines the direction of heat flow based on the actual temperature difference.

[0091] After obtaining the radiative and convective heat flux components on the surface of the freezer, the onboard computing equipment does not evaluate the impact of a single heat source in isolation. Instead, in step S4, it synthesizes the total external heat load using the principle of thermodynamic superposition. Based on the freezer's thermal resistance-capacity characteristics, it establishes a heat penetration model to predict the dynamic evolution trend of the freezer's internal temperature. This process is specifically implemented through the following sub-steps:

[0092] S41, Total external input heat flux.

[0093] The onboard computing equipment performs vector synthesis calculations, algebraically superimposing the total radiative heat flux calculated in step S2 with the total convective heat flux calculated in step S3 to obtain the total external input heat flux acting on the outer wall of the freezer at the current moment. The calculation expression is:

[0094] ;

[0095] In the formula, The total external heat flux represents the intensity of the thermal shock to the freezer from the external environment. This represents the total radiative heat flux; This represents the total convective heat flux.

[0096] In this step, the onboard computing equipment takes into account the sign and directionality of heat flux. When the external airflow temperature is lower than the surface temperature of the freezer, the convective heat flux is negative, and the convection partially cancels out the radiative heat input; conversely, the two are superimposed, exacerbating the heat load.

[0097] S42. Construct a temperature rise rate prediction model based on the lumped parameter method.

[0098] To shift from passive alarm to proactive early warning, the airborne computing equipment utilizes the lumped parameter method in heat transfer to establish a differential equation model of the temperature change inside the freezer over time. This model treats the freezer's insulation layer as a thermal resistance element and the goods and air inside the freezer as heat capacity elements.

[0099] The onboard computing equipment calculates the effective heat conduction that can penetrate the insulation layer and enter the freezer's interior. Combining this with the release capacity of any passive cooling sources (such as dry ice or refrigerant) within the freezer, it predicts the rate of temperature change inside the freezer. The prediction model satisfies the following differential equation:

[0100] ;

[0101] In the formula, The predicted rate of temperature change inside the freezer; This is the equivalent total heat capacity of the refrigerated system (including the goods); The total external input heat flux calculated in step S41; The convective heat transfer thermal resistance of the outer surface of the freezer is given by a value of ; The total thermal resistance of heat transfer from the external environment to the interior of the freezer includes the external convective thermal resistance, the thermal conductivity of the insulation layer, and the internal convective thermal resistance. The ambient temperature; The internal temperature of the freezer as measured at the current moment; This represents the total surface area of ​​the freezer. This is the rated cooling power of the cold source inside the freezer. If it is a passive freezer and no refrigerant is placed inside, this item is zero.

[0102] S43, Generate heat avoidance trigger command.

[0103] The onboard computing equipment will calculate the predicted rate of temperature change. With the preset safe temperature rise rate threshold Real-time comparison is performed. Safe temperature rise rate threshold. It is a critical value determined based on the temperature sensitivity of the goods and the remaining transportation time.

[0104] If the comparison result shows This indicates that, given the current flight attitude and environmental conditions, the internal temperature of the freezer will exceed the safety limit in a short period of time. The onboard computing equipment then generates a thermal avoidance trigger signal, activating the subsequent attitude optimization control logic.

[0105] like If the current flight status is maintained, the monitoring cycle will continue for the next period. This mechanism allows the system to identify high-risk heat input trends before the internal temperature actually exceeds the limit.

[0106] When the onboard computing equipment determines that the thermal risk faced by the refrigerator exceeds the safety threshold, it does not directly take a single evasive action. Instead, it parameterizes the physical properties of the flight space, constructs a virtual potential field that includes thermal environment constraints and flight energy consumption constraints, and solves for the minimum point in this potential field through an optimization algorithm to determine the optimal flight attitude. This process is specifically implemented through the following sub-steps:

[0107] S51. Construct anisotropic thermal impedance function.

[0108] The onboard computing device establishes the attitude angle vector of the UAV. Let be the virtual thermal impedance function, the independent variable. This function characterizes the intensity of external heat flux experienced by the freezer when the drone is in a specific attitude. Due to the asymmetric geometry of the freezer and the definite directionality of solar radiation and wind fields, this function exhibits anisotropic characteristics in three-dimensional attitude space. The mathematical expression for the thermal impedance function is constructed as follows:

[0109] ;

[0110] In the formula, for posture The corresponding thermal cost value; This is the heat cost weighting coefficient, which is set according to the temperature sensitivity level of the transported goods. For highly sensitive goods (such as vaccines and blood), this weighting is set to a larger value. In posture The predicted total external input heat flux is derived from the calculation logic of the aforementioned steps S2-S4.

[0111] In this step, the airborne computing equipment uses virtual simulation to traverse or sample the possible attitude space at the current moment and calculate different yaw angles. (e.g., backlighting, sidelighting) and different tilt angles Expected heat flux under conditions such as side-flying.

[0112] S52. Construct the flight energy consumption impedance function.

[0113] To prevent drones from engaging in high-energy-consuming flight maneuvers (such as steep side-flight or high angle-of-attack wind resistance) in an excessive pursuit of low-temperature attitudes, which could lead to battery depletion and safety incidents, the onboard computing equipment incorporates a flight energy consumption impedance function. This function estimates the motor power required to maintain a specific attitude based on the UAV's aerodynamic model. The flight energy consumption impedance function is expressed as:

[0114] ;

[0115] In the formula, for posture The corresponding energy consumption cost; This is the energy consumption weighting coefficient, which is inversely proportional to the drone's current remaining battery power. In other words, the lower the battery power, the higher the weight given to energy consumption. The aerodynamic drag coefficient is attitude-dependent and changes nonlinearly as the UAV's frontal area changes with its attitude. air density; The relative airflow velocity modulus; For reference windward area of ​​the drone; This refers to the basic hovering power of the drone.

[0116] S53. Solve for the optimal attitude saddle point in a multi-objective potential field.

[0117] The airborne computing equipment linearly weights the thermal impedance function and the energy consumption impedance function to synthesize the total potential cost function. :

[0118] ;

[0119] In the formula, for posture The corresponding energy consumption cost; for posture The corresponding thermal cost value; Let be the cost function of the total potential field.

[0120] The airborne computing equipment employs a gradient descent optimization algorithm to search for the target attitude vector that minimizes the total cost function within the constraints allowed by the attitude (e.g., maximum tilt angle limit). The iterative solution process satisfies the following gradient update law:

[0121] ;

[0122] In the formula, and The first Second and third The attitude vector of the next iteration; This is the iteration step size; This is the gradient operator for attitude variables.

[0123] Using this algorithm, the airborne computing equipment can find an equilibrium point (i.e., a potential field saddle point) that corresponds to an attitude that can reduce the heat input of the refrigerator without causing excessive increase in flight energy consumption.

[0124] S54. Generate and execute flight control commands.

[0125] After calculating the optimal attitude vector, the onboard computing equipment parses it into specific flight control commands. If the optimal attitude involves yaw angle adjustment (e.g., changing the nose direction to use the fuselage to block sunlight), the onboard computing equipment generates a yaw command and sends it to the flight controller, driving the UAV to rotate around the ZB axis to the target angle.

[0126] If the optimal attitude involves tilt angle adjustments (e.g., using a sideslip flight mode to reduce aerodynamic heating of the windward side), the onboard computing equipment generates roll or pitch commands to drive the UAV to change its attitude angle and maintain force balance in that attitude.

[0127] After receiving the command, the flight controller controls the speed of each rotor motor to smoothly transition the drone to the optimal attitude, thereby ensuring flight safety and endurance while achieving active thermal avoidance of the freezer.

[0128] Example 2:

[0129] Please see the appendix Figure 2 A freezer, comprising:

[0130] Cabinet 1 is a hollow box structure with enclosed storage space, used to store goods that require low-temperature preservation, such as vaccines and blood products. The shell wall of Cabinet 1 is not a single-layer structure; its walls are filled with an insulation layer 5.

[0131] In this embodiment, the insulation layer 5 is a composite material of aerogel felt and vacuum insulation board, which has an extremely low thermal conductivity. This is used to isolate the external thermal environment from the internal space to the greatest extent and reduce the amount of heat entering the cabinet through thermal conduction.

[0132] To achieve active temperature control, a refrigeration compressor 2 is fixedly installed inside cabinet 1. Considering the special characteristics of drone transportation scenarios, a Stirling refrigeration compressor is preferred for refrigeration compressor 2. Unlike traditional vapor compression refrigeration machines that rely on gravity-fed oil return, the Stirling refrigeration compressor utilizes the reciprocating motion of a gaseous working fluid between the expansion and compression chambers for cooling. It contains no liquid lubricating oil, thus possessing extremely strong vibration and tilt resistance. This allows the refrigeration compressor 2 to maintain continuous and stable cooling operation even under multi-angle shaking conditions caused by significant yaw and tilt changes in the drone's attitude, without experiencing cylinder jamming or cooling interruption.

[0133] The cooling output end of the refrigeration compressor 2 is connected to a refrigeration end 3, which is installed on one side of the refrigeration compressor 2 and extends along its axial direction. A refrigeration end head 4 is further connected to the end of the refrigeration end 3. In order to improve heat exchange efficiency, the refrigeration end head 4 is made of a high thermal conductivity metal such as copper, aluminum or their alloy.

[0134] In this embodiment, the transfer of cooling capacity employs a contact-type heat conduction design. Specifically, the cooling end 4 contacts the heat-conducting wall surface inside the cabinet 1 (e.g., the outer wall of the aluminum alloy inner liner of the cabinet 1). To reduce contact thermal resistance, the shape of the cooling end 4 is machined to perfectly match the contour of the contact area of ​​the cabinet 1 (e.g., both are flat surfaces or matching curved surfaces) to maximize the contact heat transfer area. Through this close physical contact, the cooling end 4 can efficiently transfer the cooling capacity generated by the refrigeration compressor 2 directly to the metal inner wall of the cabinet 1 via contact heat conduction, thereby cooling the internal space and avoiding the airflow short-circuiting problem that may occur in traditional air-cooled systems when the drone is shaking.

[0135] In addition, to achieve thermal environment sensing, temperature sensor 6 is arranged inside the insulation layer 5 or attached to the inner wall of the cabinet 1. Temperature sensor 6 is used to collect the real-time temperature inside the refrigerator and generate corresponding temperature data. This temperature data is transmitted to the onboard computing equipment of the UAV through a data interface, serving as a key feedback signal for correcting the thermal model and triggering thermal avoidance strategies.

[0136] In summary, this freezer establishes a hardware foundation for adapting to the high-dynamic flight environment of UAVs through the anti-disturbance characteristics of the refrigeration compressor 2 (Stirling model) and the efficient contact heat transfer structure of the refrigeration end 4. Together with the aforementioned intelligent monitoring methods, it achieves comprehensive thermal safety protection for goods.

Claims

1. A method for intelligent monitoring of refrigerated containers based on unmanned aerial vehicle (UAV) transportation, characterized in that, Includes the following steps: S1. Collect flight status data, environmental parameter data, and geometric and physical parameter data of the refrigerated container from the UAV; S2. Based on the Euler angles in the flight status data and the solar incident vector at the current moment, construct a geometric projection model, calculate the effective light-receiving area of ​​each surface of the freezer through the geometric projection model, and calculate the radiative heat flux acting on the surface of the freezer in combination with the solar radiation intensity. S3. Based on the instantaneous current values ​​of each rotor motor of the UAV in the flight status data, extract the high-frequency disturbance characteristics of the current, use the high-frequency disturbance characteristics of the current as the observation variable characterizing the turbulence intensity of the rotor downwash flow field, perform in-situ correction on the benchmark convective heat transfer coefficient calculated based on the relative wind speed, obtain the actual convective heat transfer coefficient of the freezer surface, and calculate the convective heat flux of the freezer surface accordingly. S4. The radiative heat flux and the convective heat flux are vectorized to obtain the total external heat flux acting on the freezer. The lumped parameter method is used to predict the rate of change of the internal temperature of the freezer within a preset future time period. S5. When the predicted rate of change of the internal temperature of the freezer exceeds a preset threshold, an anisotropic thermal impedance flight potential field containing thermal cost and flight energy consumption cost is constructed. In the anisotropic thermal impedance flight potential field, the gradient optimization algorithm is used to solve for the optimal target attitude that minimizes the total cost. Based on the optimal target attitude, flight control commands are generated and sent to the flight controller to drive the UAV to adjust the yaw angle or tilt angle to avoid high heat flux areas.

2. The intelligent monitoring method for refrigerated containers based on unmanned aerial vehicle (UAV) transportation according to claim 1, characterized in that, The flight status data includes at least the UAV's Euler angles, three-dimensional position coordinates, ground speed vector, real-time rotational speed of each rotor motor, and instantaneous phase current of each rotor motor at the current moment. The environmental parameter data includes at least ambient temperature and relative air velocity vector; The geometric and physical parameter data include at least the area of ​​each surface of the freezer, the normal vector of the outer surface, the thermal conductivity of the insulation layer, the thickness of the insulation layer, and the combined specific heat capacity of the freezer and the internal load.

3. The intelligent monitoring method for refrigerated containers based on unmanned aerial vehicle (UAV) transportation according to claim 1, characterized in that, In step S2, the calculation of the radiant heat flux acting on the surface of the freezer specifically includes the following steps: Transform the solar unit direction vector in the geodetic coordinate system to the body coordinate system to obtain the solar incident vector; Traverse each surface of the freezer and calculate the dot product of the solar incident vector and the outer surface normal vector of each surface of the freezer in the body coordinate system. The surface with a positive dot product result is selected as the sun-facing surface. The dot product value of the sun-facing surface, the physical area of ​​the surface, the solar radiation absorptivity, and the solar irradiance of the current environment are multiplied to obtain the one-sided radiative heat flux. The radiant heat flux is obtained by summing the radiant heat flux of each light-facing surface of the freezer.

4. The intelligent monitoring method for refrigerated containers based on unmanned aerial vehicle (UAV) transportation according to claim 1, characterized in that, Step S3, the extraction of high-frequency disturbance features of the current, specifically includes the following steps: Set a sampling time window and obtain the instantaneous phase current sequence of each rotor motor of the UAV within the sampling time window; Calculate the average current value of each rotor motor of the UAV within the sampling time window; Calculate the square of the difference between each data point in the instantaneous phase current sequence and the average current value, then integrate or sum and average the square values ​​and take the square root to obtain the root mean square value reflecting the severity of current fluctuations. The root mean square value is defined as the current disturbance index.

5. The intelligent monitoring method for refrigerated containers based on unmanned aerial vehicle (UAV) transportation according to claim 1, characterized in that, In step S3, the reference convective heat transfer coefficient is corrected in situ, specifically including the following steps: Based on the fluid dynamics theory of flat plate swirling flow, the reference convective heat transfer coefficient is calculated according to the Reynolds number and Prandtl number. A correction factor is constructed, which is obtained by adding the product of the flow-heat coupling weight coefficient and the normalized disturbance amount to a constant 1. The normalized disturbance amount is the ratio of the current disturbance index to the total average current of all motors. The actual convective heat transfer coefficient is obtained by multiplying the reference convective heat transfer coefficient by the correction factor.

6. The intelligent monitoring method for refrigerated containers based on unmanned aerial vehicle (UAV) transportation according to claim 1, characterized in that, In step S4, predicting the rate of change of the internal temperature of the freezer within a preset future time period specifically includes the following steps: The convective heat transfer resistance of the outer surface of the freezer is calculated based on the actual convective heat transfer coefficient and the total surface area of ​​the freezer. Combined with the thermal conductivity resistance of the insulation layer (5) and the convective heat transfer resistance of the inner surface, the total heat transfer resistance from the external environment to the interior of the freezer is calculated by series superposition. Calculate the effective transmission component of the total external heat flux under the influence of the total heat transfer resistance, and the temperature difference heat transfer component of the difference between the ambient temperature and the internal temperature under the influence of the total heat transfer resistance. The net heat input is obtained by subtracting the rated cooling power of the cold source inside the freezer from the sum of the effective transmission component and the temperature difference heat transfer component. The rate of change is obtained by dividing the net heat input by the combined specific heat capacity of the freezer and the load of goods inside.

7. The intelligent monitoring method for refrigerated containers based on unmanned aerial vehicle (UAV) transportation according to claim 1, characterized in that, In step S5, the construction of the anisotropic thermal impedance flight potential field specifically includes the following steps: A thermal impedance function is established with the UAV attitude angle vector as the independent variable. The value of the thermal impedance function is the product of the thermal cost weighting coefficient and the predicted value of the total external heat flux under the corresponding attitude. A flight energy consumption impedance function is established with the UAV attitude angle vector as the independent variable. The value of the flight energy consumption impedance function is the product of the energy consumption weighting coefficient and the estimated total flight power. The estimated total flight power is obtained by combining the UAV attitude angle vector with the aerodynamic model. The estimated total flight power includes the basic hovering power and aerodynamic power. The aerodynamic power is proportional to the air density, the square of the relative air velocity, the reference frontal area, and the aerodynamic drag coefficient that changes with attitude. The thermal impedance function is added to the flight energy consumption impedance function to form the total potential cost function.

8. The intelligent monitoring method for refrigerated containers based on unmanned aerial vehicle (UAV) transportation according to claim 1, characterized in that, Step S5, solving for the optimal target pose, specifically includes the following steps: Initialize the attitude vector and set the iteration step size; Calculate the gradient vector of the total potential cost function with respect to the attitude angle variables; The attitude vector is updated in the negative direction of the gradient, so that the UAV's attitude evolves in the direction of the fastest decrease of the total potential cost function, until the magnitude of the gradient vector is less than the convergence threshold. The converged attitude vector is then determined as the optimal target attitude.

9. A freezer, characterized in that, The method for intelligent monitoring of refrigerated containers based on unmanned aerial vehicle (UAV) transportation, as described in any one of claims 1-8, comprises: The cabinet (1) is constructed as a hollow box structure with a closed storage space. The wall of the cabinet (1) is filled with a heat insulation layer (5) to isolate the external thermal environment from the internal storage space. A refrigeration compressor (2) is fixedly installed inside the cabinet (1) and is used to generate a low-temperature cold source required for refrigeration. The refrigeration end (3) is installed on one side of the refrigeration compressor (2) and extends along the refrigeration compressor (2); The refrigeration end (4) is connected to the end of the refrigeration end (3) and abuts against the heat-conducting wall of the cabinet (1) to transfer the cold energy generated by the refrigeration compressor (2) to the inside of the cabinet (1) by means of contact heat conduction. A temperature sensor (6) is arranged inside the insulation layer (5) to collect the real-time temperature inside the freezer and generate corresponding temperature data.

10. A freezer according to claim 9, characterized in that, The refrigeration compressor (2) is a Stirling refrigeration compressor, which is used to maintain continuous refrigeration in the multi-angle shaking environment caused by the change of the attitude of the UAV. The cooling end (4) is made of a high thermal conductivity metal material, and its shape is adapted to the contact part of the cabinet (1) to maximize the contact heat transfer area.

Images