Method, device and equipment for optimizing parameters affecting carbon emissions of cold chain delivery vehicles

Abstract

The present invention provides a method, device, electronic device and storage medium for optimizing parameters affecting carbon emissions of cold chain delivery vehicles, including: obtaining speed parameters, refrigeration parameters, cargo weight and road condition parameters of the cold chain delivery vehicle during driving; while keeping cargo weight and road condition parameters unchanged, based on the change threshold of the speed parameter and the change threshold of the refrigeration parameter, changing the speed parameter value and the refrigeration parameter value to generate multiple parameter value combinations, the parameter value combinations include speed parameters, refrigeration parameters, cargo weight and road condition parameters; based on the carbon emission prediction model of the cold chain delivery vehicle, predicting the carbon emissions corresponding to each parameter value combination, and outputting the speed parameter value and the refrigeration parameter value corresponding to the parameter value combination with the minimum carbon emissions. This solves the problem that the prior art cannot achieve the optimization and regulation of carbon emission influencing factors of cold chain delivery vehicles based on carbon emissions.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle carbon emission control, and in particular to a method, device and equipment for optimizing parameters affecting carbon emissions of a cold chain delivery vehicle. Background Art

[0002] Carbon dioxide can cause the greenhouse effect, and the main source of carbon dioxide emissions in daily life is carbon emissions from vehicles. And with the introduction of the dual carbon goals, the cold chain distribution link, which needs to maintain the dual energy needs of driving power and low temperature environment, needs to control carbon dioxide emissions.

[0003] At present, the calculation of transportation carbon emissions mainly calculates regional transportation carbon emissions from a macro level, mainly based on three parameters: average carbon emissions of different vehicles, number of motor vehicles, and average driving distance of different vehicles. However, cold chain delivery vehicles are different from other vehicles. They have dual requirements for driving power and refrigeration environment. The existing technology cannot calculate the carbon emissions of cold chain delivery vehicles, and it is also impossible to optimize and control the factors affecting vehicle carbon emissions based on carbon emissions. Summary of the invention

[0004] In response to the problems existing in the prior art, the present invention provides a method, device and equipment for optimizing parameters influencing the carbon emissions of cold-chain delivery vehicles, so as to solve the problem that the prior art cannot achieve optimal regulation of factors influencing vehicle carbon emissions based on carbon emissions, and realize the optimal regulation of carbon emissions of cold-chain delivery vehicles, thereby achieving the technical effect of reducing the carbon emissions of cold-chain delivery vehicles.

[0005] In a first aspect, the present invention provides a method for optimizing parameters affecting carbon emissions of cold-chain delivery vehicles, comprising: obtaining speed parameters, refrigeration parameters, cargo weight and road condition parameters of the cold-chain delivery vehicle during driving; while keeping the cargo weight and the road condition parameters unchanged, based on a change threshold of the speed parameter and a change threshold of the refrigeration parameter, changing the speed parameter value and the refrigeration parameter value to generate multiple parameter value combinations, wherein the parameter value combinations include speed parameters, refrigeration parameters, cargo weight and road condition parameters; predicting the carbon emissions corresponding to each parameter value combination based on a carbon emission prediction model for cold-chain delivery vehicles, and outputting the speed parameter value and refrigeration parameter value corresponding to the parameter value combination with the smallest carbon emissions; wherein the carbon emission prediction model for cold-chain delivery vehicles is trained based on sample parameter value combinations with carbon emission labels.

[0006] Optionally, the road condition parameters of the cold chain delivery vehicle during driving are obtained in the following manner, including: obtaining a road condition image of the cold chain delivery vehicle during driving; identifying the road condition image through a road condition information recognition model, and determining the number of vehicles, vehicle categories, and pixels of a vehicle prediction box in the road condition image; determining the pixel area ratio of the vehicle based on the number of vehicles, the vehicle category, and the pixels of the vehicle prediction box, and the pixel area of ​​the vehicle is the road condition parameter.

[0007] Optionally, the pixel area ratio of the vehicle is determined based on the number of vehicles, the vehicle category, and the pixels of the vehicle prediction box, and is implemented by the following formula: Among them, i represents the vehicle category, S i represents the pixel area ratio of the i-th type of vehicle, n i is the number of vehicles of the i-th category in the road condition image, W ij is the horizontal pixel of the jth vehicle prediction box in the i-th vehicle, H ij is the vertical pixel of the prediction box of the jth vehicle in the i-th vehicle category, and Z is the pixel area of ​​the road condition image.

[0008] Optionally, the cold chain delivery vehicle carbon emission prediction model is obtained on the basis of the iTransformer model by replacing the layer normalization equation with the depth normalization equation, and replacing the multivariate attention mechanism with a sliding window attention mechanism that integrates global information.

[0009] Optionally, the road condition information recognition model is obtained by introducing full-dimensional dynamic convolution into the convolution layer for deep feature extraction in the backbone network based on the YOLOv8s model, and adding a progressive feature pyramid network between the convolution layer in the neck network and the convolution layer for deep feature extraction and between two convolution modules for deep feature extraction.

[0010] Optionally, the refrigeration parameters include refrigeration temperature, refrigeration humidity and refrigeration power.

[0011] In a second aspect, the present invention provides an optimization device for parameters affecting carbon emissions of cold-chain delivery vehicles, comprising: an acquisition module for acquiring speed parameters, refrigeration parameters, cargo weight and road condition parameters of the cold-chain delivery vehicle during driving; a generation module for changing the speed parameter value and the refrigeration parameter value to generate multiple parameter value combinations based on a change threshold of the speed parameter and a change threshold of the refrigeration parameter while keeping the cargo weight and the road condition parameters unchanged, wherein the parameters included in the parameter value combination are speed parameter, refrigeration parameter, cargo weight and road condition parameter; a first processing module for predicting the carbon emissions corresponding to each parameter value combination based on a carbon emission prediction model for cold-chain delivery vehicles, and outputting the speed parameter value and refrigeration parameter value corresponding to the parameter value combination with the smallest carbon emissions; wherein the carbon emission prediction model for cold-chain delivery vehicles is trained based on sample parameter value combinations with carbon emission labels.

[0012] In a third aspect, the present invention further provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, it implements a method for optimizing parameters affecting the carbon emissions of a cold chain delivery vehicle as described in any one of the above-mentioned methods.

[0013] In a fourth aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a method for optimizing parameters affecting the carbon emissions of a cold chain delivery vehicle as described in any of the above.

[0014] In a fifth aspect, the present invention further provides a computer program product, comprising a computer program, which, when executed by a processor, implements a method for optimizing parameters affecting the carbon emissions of a cold chain delivery vehicle as described in any of the above-mentioned methods.

[0015] The method for optimizing the parameters influencing the carbon emissions of a cold-chain delivery vehicle provided by the present invention obtains the speed parameters, refrigeration parameters, cargo weight and road condition parameters of the cold-chain delivery vehicle during driving; while keeping the cargo weight and the road condition parameters unchanged, within the change threshold range of the speed parameters and the refrigeration parameters, the speed parameter values ​​and the refrigeration parameter values ​​are changed to generate multiple parameter value combinations, the carbon emissions corresponding to each of the parameter value combinations are determined, and the speed parameter values ​​and the refrigeration parameter values ​​corresponding to the parameter value combination with the smallest carbon emissions are output. The speed parameters and the refrigeration parameters corresponding to the parameter combination with less carbon emissions are determined through the carbon emissions corresponding to the current parameters influencing the carbon emissions of the cold-chain delivery vehicle, thereby solving the problem in the prior art that it is impossible to optimize and regulate the factors influencing the carbon emissions of the cold-chain delivery vehicle based on the carbon emissions, and achieving the technical effect of reducing the carbon emissions of the cold-chain delivery vehicle by dynamically adjusting the influencing parameters, thereby achieving energy conservation and emission reduction. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0017] Figure 1 It is a flow chart of a method for optimizing parameters affecting carbon emissions of cold chain delivery vehicles provided by the present invention.

[0018] Figure 2 It is a structural schematic diagram of the road condition information recognition model provided by the present invention.

[0019] Figure 3 It is a schematic diagram of the structure of the full-dimensional dynamic convolution provided by the present invention.

[0020] Figure 4 It is a schematic diagram of the structure of the progressive feature pyramid network provided by the present invention.

[0021] Figure 5 It is a structural schematic diagram of the dynamic prediction model of carbon emissions from cold chain distribution under multi-source information fusion of iTransformer-GSD provided by the present invention.

[0022] Figure 6 It is a structural schematic diagram of a device for optimizing parameters affecting carbon emissions of a cold chain delivery vehicle provided by the present invention.

[0023] Figure 7 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0024] In the present invention, the term "and / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B may represent: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the associated objects before and after are in an "or" relationship.

[0025] In the present invention, the term "plurality" refers to two or more than two, and other quantifiers are similar to them.

[0026] In the present invention, the terms "first", "second", etc. are used to distinguish similar objects, and are not used to describe a specific order or sequence. It should be understood that the terms used in this way can be interchangeable under appropriate circumstances, so that the embodiments of the present invention can be implemented in an order other than those illustrated or described herein, and the objects distinguished by "first" and "second" are generally of the same type, and the number of objects is not limited. For example, the first object can be one or more.

[0027] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. 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.

[0028] Figure 1 FIG. 1 is a flow chart of a method for optimizing parameters affecting carbon emissions of cold chain delivery vehicles provided by the present invention. Figure 1 As shown, the method includes the following steps S102-S106:

[0029] Step S102: Obtain speed parameters, refrigeration parameters, cargo weight, and road condition parameters of the cold chain delivery vehicle during driving.

[0030] In an exemplary embodiment, the refrigeration parameters include refrigeration temperature, refrigeration humidity and refrigeration power.

[0031] It should be noted that lower refrigeration temperatures require the refrigeration system to work continuously to maintain the required temperature, resulting in higher consumption of fuel or electricity, thereby increasing carbon emissions. Similarly, in order to control humidity, cold chain delivery vehicles may require additional equipment or energy to maintain the required humidity level, which will also lead to an increase in carbon emissions.

[0032] As an optional embodiment, the speed parameters of the cold chain delivery vehicle during driving can be collected by a driving recorder.

[0033] As an optional embodiment, the refrigeration parameters can be recorded by the control unit to obtain the power of the refrigeration unit every second, and the temperature and humidity sensor can be used to collect the changes in temperature and humidity in the cold chain delivery vehicle compartment every second.

[0034] As an optional embodiment, the cargo weight can be counted in the cold chain delivery vehicle compartment through a transport record sheet. For unmanned cold chain delivery vehicles, the cargo weight can also be directly obtained through a pressure sensor in the compartment.

[0035] In an exemplary embodiment, the road condition parameters of the cold chain delivery vehicle during driving can be obtained through the following steps S11-S13:

[0036] Step S11: Acquire a road condition image of the cold chain delivery vehicle during driving.

[0037] Step S12: Identify the road condition image using a road condition information recognition model to determine the number of vehicles, vehicle categories, and pixels of a vehicle prediction box in the road condition image.

[0038] In an exemplary embodiment, the road condition information recognition model is obtained by introducing full-dimensional dynamic convolution into the convolution layer for deep feature extraction in the backbone network based on the YOLOv8s model, and adding a progressive feature pyramid network between the convolution layer in the neck network and the convolution layer for deep feature extraction and between two convolution modules for deep feature extraction.

[0039] As an optional embodiment, Figure 2 It is a structural diagram of the road condition information recognition model provided by the present invention. The road condition information recognition model in the present invention is YOLOv8s-OA obtained by improving the YOLOv8s model. As shown in the figure, the road condition information recognition model includes a neck network, a backbone network and a detection head. By introducing the full-dimensional dynamic convolution (Omni-Dimensional Dynamic Convolution, referred to as ODConv) in the convolution layer for deep feature extraction in the backbone network, that is, C2f-ODC in the figure. And add a progressive feature pyramid network (Asymptotic Feature Pyramid Network, referred to as AFPN) between the convolution layer in the neck network and the convolution layer for deep feature extraction, and between the two convolution modules for deep feature extraction, that is Figure 2 ASF-2 and ASF-3 in the road condition information recognition model. It should be noted that the convolution layer Conv in the road condition information recognition model is used to extract the local spatial features of the object; C2f-ODC is used to extract features and extract vehicle information from driving images; SPPF is a pooling layer used to fuse local features and global features; C2f is used to extract features; ASF is used to assign weights to features at different levels to enhance the importance of key features; the detection head Detect is used to extract the category and location information of the object to achieve object classification and positioning.

[0040] As an optional embodiment, Figure 3 Schematic diagram of the structure of the full-dimensional dynamic convolution (i.e., ODConv) provided by the present invention, wherein GAP is used for global average pooling, FC is a fully connected layer, and ReLU and Sigmoid are both activation functions.si Represents the attention value of the convolution kernel in time and space, a ci represents the input channel attention value, a fi represents the output channel attention value, a wi Represents the overall convolution kernel attention value, w i Represents the i-th convolution kernel, and N represents the number of convolution kernels. Full-dimensional dynamic convolution helps the model to locate more accurately and improve the detection performance of the model by learning complementary attention in four dimensions of the filter space.

[0041] As an optional embodiment, Figure 4 The ASF is a schematic diagram of the structure of the progressive feature pyramid network provided by the present invention. The ASF is used to assign weights to features at different levels and enhance the importance of key features. By narrowing the semantic gap between non-adjacent scales, the semantic information in the high level and the detail information in the low level are fully integrated to avoid information loss in multi-level transmission.

[0042] It should be noted that through the above improvements, the road condition information recognition model can learn the differences in different dimensions, fully capture the key context information, improve the feature extraction capability, avoid the loss of key information, and thus improve the accuracy of vehicle recognition.

[0043] Step S13: Determine the pixel area ratio of the vehicle based on the number of vehicles, the vehicle category and the pixels of the vehicle prediction frame, where the pixel area of ​​the vehicle is the road condition parameter.

[0044] In an exemplary embodiment, the pixel area ratio of the vehicle is determined based on the number of vehicles, the vehicle category, and the pixels of the vehicle prediction box, which can be obtained by the following formula:

[0045]

[0046] It should be noted that, i represents the vehicle category, S i represents the pixel area ratio of the i-th type of vehicle, n i is the number of vehicles of the i-th category in the road condition image, W ij is the horizontal pixel of the jth vehicle prediction box in the i-th vehicle, H ij is the vertical pixel of the prediction box of the jth vehicle in the i-th vehicle category, and Z is the pixel area of ​​the road condition image.

[0047] As an optional embodiment, the vehicle category may be cars, trucks, and buses. That is, the road condition parameters include the pixel area ratio of cars and the number of cars, the pixel area ratio of trucks and the number of trucks, and the pixel area ratio of buses and the number of buses. The number of vehicles of different categories and their respective pixel area ratios are used to characterize the congestion of the road condition.

[0048] It should be noted that when the road conditions are poor, congested roads will cause vehicles to idle or drive slowly for a long time, which will increase the vehicle's carbon emissions. When the vehicle is idling, the engine fuel efficiency is low, and the carbon emissions generated per unit distance traveled will also increase. When the road conditions are complex, the driver needs to brake or slow down frequently, which will also lead to fuel waste and increase carbon emissions. Therefore, road condition parameters need to be taken into consideration when calculating carbon emissions.

[0049] Step S104: While keeping the cargo weight and the road condition parameters unchanged, based on the change threshold of the speed parameter and the change threshold of the refrigeration parameter, change the speed parameter value and the refrigeration parameter value to generate multiple parameter value combinations, wherein the parameter value combinations include the speed parameter, the refrigeration parameter, the cargo weight and the road condition parameter.

[0050] As an optional embodiment, the speed parameter change threshold and the cooling parameter change threshold described in the present invention can be limited by specific numerical values ​​or by percentages, and the speed parameter change threshold and the cooling parameter change threshold are usually different. The specific threshold range can be flexibly set, and the embodiment of the present application does not further limit this.

[0051] As an optional embodiment, assuming that the current speed of the cold chain delivery vehicle is 60km / h, the refrigeration parameters are: temperature 2 degrees Celsius, humidity 85%, refrigeration power 7 kilowatts (KW), the speed parameter change threshold is ±20, the temperature change threshold is ±2, the humidity change threshold is ±5%, and the refrigeration power threshold is ±2KW. The speed parameter range is 40-80 (km / h), the temperature range is 0-4 (degrees Celsius), the humidity range is 75%-95%, and the refrigeration power range is 5-9 (KW). The cargo weight is 9000kg, and the road condition parameters are: 6 cars, S 轿车 =60%; 4 trucks, S 货车 = 20% and 2 buses, S 公交车 =20%. Then a possible combination of parameter values ​​generated is shown in Table 1 below:

[0052] Table 1

[0053]

[0054] It should be noted that the value in the above table is only one possibility. In actual application, the value interval of the parameter value combination generated can be as small as possible so that the generated parameter value combination can cover more possibilities.

[0055] Step S106: Based on the carbon emission prediction model for cold chain delivery vehicles, the carbon emissions corresponding to each parameter value combination are predicted, and the speed parameter value and the refrigeration parameter value corresponding to the parameter value combination with the minimum carbon emissions are output.

[0056] It should be noted that the cold chain delivery vehicle carbon emission prediction model is obtained by training based on a combination of sample parameter values ​​with carbon emission labels.

[0057] As an optional embodiment, the sample parameter value combination can be divided into a training set, a validation set and a test set in a ratio of 8:1:1 to achieve better training for the carbon emission prediction model of cold chain delivery vehicles.

[0058] As an optional embodiment, it is assumed that the carbon emissions of the above combination are as shown in Table 2 below:

[0059] Table 2

[0060]

[0061] From the above table, we can see that the minimum value of carbon emission is 125, and the corresponding speed parameter is 70 (km / h) and the corresponding cooling parameters are: temperature 3 degrees Celsius, humidity 90%, and cooling power 8KW.

[0062] As an optional embodiment, the output prompt information can be output in the form of voice to prompt the driver to adjust the vehicle speed to 70, the cooling parameters to the temperature of 3 and the humidity of 52%, thereby achieving the technical effect of reducing the carbon emissions of the vehicle.

[0063] As an optional embodiment, for unmanned cold chain delivery vehicles, there is no need to output prompt information, and the original speed parameters and refrigeration parameters can be directly adjusted to the speed parameter values ​​and refrigeration parameter values ​​corresponding to the parameter value combination with the lowest carbon emissions.

[0064] In an exemplary embodiment, the cold chain delivery vehicle carbon emission prediction model is obtained on the basis of the iTransformer model by replacing the layer normalization equation with the depth normalization equation and replacing the multivariate attention mechanism with a sliding window attention mechanism that integrates global information.

[0065] As an optional embodiment, the present invention introduces a sliding window attention mechanism (Global+Sliding Window Attention) and DeepNorm (i.e., the above-mentioned deep normalization equation) that integrates global information on the basis of the iTransformer model, thereby reducing the computational complexity and improving the prediction accuracy of carbon emissions for refrigerated vehicles, thereby obtaining a dynamic prediction model for carbon emissions of cold chain distribution under multi-source information fusion of iTransformer-GSD. Figure 5 It is a structural schematic diagram of the dynamic prediction model of carbon emissions from cold chain distribution under multi-source information fusion of iTransformer-GSD provided by the present invention.

[0066] It should be noted that iTransformer-GSD maps the entire time series data of carbon emission input features into high-dimensional feature representation, and reflects its historical change process with variables as the main body. First, the sliding window attention mechanism that integrates global information considers the importance of local context, such as Figure 5 As shown in (b), a fixed-size window is used to collect correlation around the carbon emission time series data position, and global attention is added to important input positions. By stacking multiple layers of sliding window attention, a larger receptive field is generated, so that all input data positions can be accessed and representation information containing the entire input sequence is established. This makes the model computational complexity linearly proportional to the input data, greatly reducing the amount of model computation. Then, the deep normalization equation is introduced to adjust the weights of the key-value pairs in the attention layer (i.e. Figure 5 The stability and prediction accuracy of the algorithm are improved in part d). The feedforward network is used to extract the time features shared by historical observations and future predictions. Finally, the extracted features are extrapolated into carbon emission prediction results through the projection layer.

[0067] Through the above steps S102-S106, by obtaining the speed parameters, refrigeration parameters, cargo weight and road condition parameters of the cold-chain delivery vehicle during driving; while keeping the cargo weight and the road condition parameters unchanged, within the change threshold range of the speed parameter and the refrigeration parameter, change the speed parameter value and the refrigeration parameter value to generate multiple parameter value combinations, determine the carbon emissions corresponding to each parameter value combination, and output the speed parameter value and the refrigeration parameter value corresponding to the parameter value combination with the smallest carbon emissions. Through the carbon emissions corresponding to the parameters affecting the carbon emissions of the current cold-chain delivery vehicle, the speed parameters and the refrigeration parameters corresponding to the parameter combination with less carbon emissions are determined, which solves the problem that the prior art cannot achieve optimal regulation of the factors affecting the carbon emissions of the cold-chain delivery vehicle based on the carbon emissions, and achieves the technical effect of reducing the carbon emissions of the cold-chain delivery vehicle by dynamically adjusting the influencing parameters, thereby achieving energy conservation and emission reduction.

[0068] The executor of each step in the method can be an optimization device for the parameters affecting the carbon emissions of cold chain delivery vehicles. The device can be implemented through software and / or hardware. The device can be integrated in an electronic device. The electronic device can be a terminal device (such as a smart phone, a personal computer, etc.), or a server (such as a local server or a cloud server, or a server cluster, etc.), or a processor, or a chip, etc.

[0069] Obviously, the above-described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. In order to better understand the above method, the above process is described below in conjunction with the embodiments, but the technical solutions of the embodiments of the present invention are not limited. Specifically:

[0070] The present invention constructs a carbon emission data non-sensing collection system based on the factors affecting the carbon emissions of cold-chain delivery vehicles (i.e., the speed parameters, refrigeration parameters, cargo weight, and road condition parameters in the above-mentioned embodiments) to realize the collection of carbon emission data of refrigerated delivery vehicles. And with the carbon emission influencing factors as input, a carbon emission prediction model for cold-chain delivery vehicles based on the improved deep learning model is constructed. Through the carbon emission prediction model for cold-chain delivery vehicles, accurate prediction of carbon emissions during the delivery process of cold-chain delivery vehicles is achieved, and the control values ​​(corresponding to the speed parameter values ​​and refrigeration parameter values ​​in the above-mentioned embodiments) are outputted, thereby realizing the optimal control of carbon emissions of cold-chain delivery vehicles.

[0071] 1. Collection of speed parameters, refrigeration parameters, cargo weight and road condition parameters

[0072] The speed parameter can be collected by a driving recorder to record the driving speed of the cold chain delivery vehicle per second, and the control unit can be used to record the power of the refrigeration unit per second. The refrigeration parameters can be collected by a temperature and humidity sensor, including the changes in temperature and humidity per second in the cold chain delivery compartment, and the weight of the goods in the cold chain delivery compartment can be counted using a transportation record table. When determining the road condition parameters, the road condition video of the cold chain delivery vehicle during driving can be collected by a driving recorder, and a frame can be randomly selected from each second of the road condition video as a road condition image. In the present invention, the YOLOv8s-OA road condition information recognition model is used to detect vehicles in the road condition image. Figure 2 It is a structural schematic diagram of the road condition information recognition model provided by the present invention.

[0073] It should be noted that the YOLOv8s-OA road condition information recognition model is based on YOLOv8s as the benchmark model, and is obtained by introducing Omni-Dimensional Dynamic Convolution (ODConv) and Asymptotic Feature Pyramid Network (AFPN). Figure 3It is a schematic diagram of the structure of the full-dimensional dynamic convolution provided by the present invention. Figure 4 It is a structural diagram of the progressive feature pyramid network provided by the present invention. The YOLOv8s-OA road condition information recognition model learns complementary attention in four dimensions of the filter space through ODConv, which is conducive to the model to locate more accurately and improve the detection performance of the model. AFPN is also used to narrow the semantic gap between non-adjacent scales, fully integrate the semantic information in the high level and the detail information in the low level, and avoid information loss in multi-level transmission. Through these two improvements, the model learns the differences in different dimensions, fully captures the key context information, improves the feature extraction capability, avoids the loss of key information, and improves the accuracy of vehicle recognition.

[0074] The YOLOv8s-OA road condition information recognition model is used to detect vehicles in road condition images, count the number of different types of vehicles in the image, and calculate the vehicle pixel area ratio based on the vehicle prediction box size. The vehicle pixel area ratio calculation formula is as follows:

[0075]

[0076] Among them, i represents the vehicle category, S i represents the pixel area ratio of the i-th type of vehicle, n i is the number of vehicles of the i-th category in the road condition image, W ij is the horizontal pixel of the jth vehicle prediction box in the i-th vehicle, H ij is the vertical pixel of the prediction box of the jth vehicle in the i-th vehicle category, and Z is the pixel area of ​​the road condition image.

[0077] 2. Carbon emission prediction model for cold chain delivery vehicles

[0078] When training the carbon emission prediction model for cold chain delivery vehicles, the carbon dioxide content in the exhaust gas of cold chain delivery vehicles is collected every second through the carbon dioxide online detector installed on the cold chain delivery vehicles. The speed parameters, refrigeration parameters, cargo weight and road condition parameters obtained are used as the input of the carbon emission prediction model for cold chain delivery vehicles, and the carbon dioxide content in the exhaust gas of cold chain delivery vehicles is used as the output of the carbon emission prediction model for cold chain delivery vehicles to train iTransformer-GSD.

[0079] It should be noted that Figure 5It is a structural diagram of the dynamic prediction model of carbon emissions from cold chain distribution under multi-source information fusion of iTransformer-GSD provided by the present invention. The dynamic prediction model of carbon emissions from cold chain distribution under multi-source information fusion of iTransformer-GSD is obtained on the basis of the iTransformer model by replacing the layer normalization equation with the depth normalization equation and replacing the multivariate attention mechanism with the sliding window attention mechanism that integrates global information, thereby reducing the computational complexity and improving the prediction accuracy of carbon emissions of cold chain distribution vehicles.

[0080] Specifically, iTransformer-GSD maps the entire time series data of the carbon emission input feature into a high-dimensional feature representation, and uses variables as the main body to reflect its historical change process. First, the sliding window attention mechanism that integrates global information considers the importance of local context, uses a fixed-size window to collect relevance around the carbon emission time series data position, and adds global attention to important input positions. By stacking multiple layers of sliding window attention, a larger receptive field is generated, so that all input data positions can be accessed, and representation information containing the entire input sequence is established. This makes the model computational complexity linearly proportional to the input data, greatly reducing the amount of model computation. Then, the DeepNorm normalization equation is introduced to adjust the weights of the key-value pairs in the attention layer, such as Figure 5 (d) in the figure improves the stability and prediction accuracy of the algorithm. The feedforward network (FeedForward) is used to extract the time features shared by historical observations and future predictions. Finally, the extracted features are extrapolated into the prediction results of carbon emissions through the projection layer (Projection).

[0081] In short, the iTransforme-GSD cold chain distribution carbon emission dynamic prediction model is trained using the carbon emission training data set, and the trained carbon emission prediction model is verified using the verification data set. The model parameters are repeatedly adjusted during the training and verification process to improve the model's prediction accuracy for carbon emissions.

[0082] 3. Optimal control of carbon emissions from cold chain delivery vehicles

[0083] During actual driving, the current carbon emissions are calculated based on the current speed parameters, refrigeration parameters, cargo weight and road condition parameters of the cold chain delivery vehicle. While keeping the cargo weight and the road condition parameters unchanged, the value ranges of the current speed parameters and refrigeration parameters are determined according to preset thresholds. Within the value range, the specific values ​​of the speed parameters and the specific values ​​of the refrigeration parameters are changed to generate multiple parameter value combinations, and the carbon emissions corresponding to each parameter combination are determined. The speed parameter values ​​and the refrigeration parameter values ​​corresponding to the parameter combination with the smallest carbon emissions are output, and the driver is prompted to adjust the refrigeration conditions and speed according to the current prompt information.

[0084] The following is a description of the device for optimizing the parameters affecting the carbon emissions of a cold-chain delivery vehicle provided by the present invention. The device for optimizing the parameters affecting the carbon emissions of a cold-chain delivery vehicle described below and the method for optimizing the parameters affecting the carbon emissions of a cold-chain delivery vehicle described above can be referenced to each other.

[0085] Figure 6 Schematic diagram of the structure of the device for optimizing the parameters affecting the carbon emissions of cold chain delivery vehicles provided by the present invention. Figure 6 As shown, the device comprises:

[0086] An acquisition module 602 is used to acquire speed parameters, refrigeration parameters, cargo weight, and road condition parameters of the cold chain delivery vehicle during driving;

[0087] A generation module 604 is used to change the speed parameter value and the refrigeration parameter value to generate multiple parameter value combinations based on the change threshold of the speed parameter and the change threshold of the refrigeration parameter while keeping the cargo weight and the road condition parameters unchanged, wherein the parameter value combination includes parameters such as speed parameter, refrigeration parameter, cargo weight and road condition parameter; a first processing module 606 is used to predict the carbon emissions corresponding to each parameter value combination based on the carbon emission prediction model for cold chain delivery vehicles, and output the speed parameter value and refrigeration parameter value corresponding to the parameter value combination with the smallest carbon emissions; wherein the carbon emission prediction model for cold chain delivery vehicles is trained based on sample parameter value combinations with carbon emission labels.

[0088] Through the above-mentioned device for optimizing the parameters affecting the carbon emissions of the cold-chain delivery vehicle, the speed parameters, refrigeration parameters, cargo weight and road condition parameters of the cold-chain delivery vehicle during driving are obtained; while keeping the cargo weight and the road condition parameters unchanged, within the change threshold range of the speed parameter and the refrigeration parameter, the speed parameter value and the refrigeration parameter value are changed to generate a plurality of parameter value combinations, the carbon emissions corresponding to each of the parameter value combinations are determined, and the speed parameter value and the refrigeration parameter value corresponding to the parameter value combination with the smallest carbon emissions are output. Through the carbon emissions corresponding to the current parameters affecting the carbon emissions of the cold-chain delivery vehicle, the speed parameters and the refrigeration parameters corresponding to the parameter combination with less carbon emissions are determined, thereby solving the problem in the prior art that the carbon emission influencing factors of the cold-chain delivery vehicle cannot be optimized and regulated based on the carbon emissions, and achieving the technical effect of reducing the carbon emissions of the cold-chain delivery vehicle by dynamically adjusting the influencing parameters, thereby achieving energy conservation and emission reduction.

[0089] Optionally, the device also includes a second processing module, which is used to obtain the road condition parameters of the delivery vehicle during driving, including: obtaining a road condition image of the cold chain delivery vehicle during driving; identifying the road condition image through a road condition information recognition model, and determining the number of vehicles, vehicle categories, and pixels of the vehicle prediction box in the road condition image; based on the number of vehicles, the vehicle category, and the pixels of the vehicle prediction box, determining the pixel area ratio of the vehicle, the pixel area of ​​the vehicle is the road condition parameter.

[0090] Optionally, the second processing module is further used to determine the pixel area ratio of the vehicle by the following formula: Among them, i represents the vehicle category, S i represents the pixel area ratio of the i-th type of vehicle, n i is the number of vehicles of the i-th category in the road condition image, W ij is the horizontal pixel of the jth vehicle prediction box in the i-th vehicle, H ij is the vertical pixel of the prediction box of the jth vehicle in the i-th vehicle category, and Z is the pixel area of ​​the road condition image.

[0091] Optionally, the cold chain delivery vehicle carbon emission prediction model is obtained on the basis of the iTransformer model by replacing the layer normalization equation with the depth normalization equation, and replacing the multivariate attention mechanism with a sliding window attention mechanism that integrates global information.

[0092] Optionally, the road condition information recognition model is obtained by introducing full-dimensional dynamic convolution into the convolution layer for deep feature extraction in the backbone network based on the YOLOv8s model, and adding a progressive feature pyramid network between the convolution layer in the neck network and the convolution layer for deep feature extraction and between two convolution modules for deep feature extraction.

[0093] Optionally, the refrigeration parameters include refrigeration temperature, refrigeration humidity and refrigeration power.

[0094] It should be noted here that the above-mentioned device provided by the present invention can implement all the method steps implemented by the above-mentioned method embodiment, and can achieve the same technical effect. The parts and beneficial effects that are the same as the method embodiment in this embodiment will not be described in detail here.

[0095] Figure 7 An example of a physical structure diagram of an electronic device is shown in FIG. Figure 7As shown, the electronic device may include: a processor 710, a communications interface 720, a memory 730 and a communication bus 740, wherein the processor 710, the communications interface 720 and the memory 730 communicate with each other through the communication bus 740. The processor 710 may call the logic instructions in the memory 730 to execute the method for optimizing the carbon emission influencing parameters of the cold chain delivery vehicle, the method comprising:

[0096] Obtain speed parameters, refrigeration parameters, cargo weight, and road condition parameters of cold chain delivery vehicles during driving;

[0097] Under the condition that the cargo weight and the road condition parameter remain unchanged, based on the change threshold of the speed parameter and the change threshold of the cooling parameter, the speed parameter value and the cooling parameter value are changed to generate a plurality of parameter value combinations, wherein the parameters included in the parameter value combination are the speed parameter, the cooling parameter, the cargo weight and the road condition parameter;

[0098] Based on the carbon emission prediction model of cold chain delivery vehicles, the carbon emissions corresponding to each parameter value combination are predicted, and the speed parameter value and the refrigeration parameter value corresponding to the parameter value combination with the minimum carbon emissions are output;

[0099] Among them, the carbon emission prediction model for cold chain delivery vehicles is obtained by training based on a combination of sample parameter values ​​with carbon emission labels.

[0100] In addition, the logic instructions in the above-mentioned memory 730 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a 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, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0101] On the other hand, the present invention further provides a computer program product, the computer program product includes a computer program, the computer program can be stored in a non-transitory computer-readable storage medium, when the computer program is executed by a processor, the computer can execute the method for optimizing the carbon emission influencing parameters of the cold chain delivery vehicle provided by the above methods, the method comprising: obtaining speed parameters, refrigeration parameters, cargo weight and road condition parameters of the cold chain delivery vehicle during driving;

[0102] Under the condition that the cargo weight and the road condition parameter remain unchanged, based on the change threshold of the speed parameter and the change threshold of the cooling parameter, the speed parameter value and the cooling parameter value are changed to generate a plurality of parameter value combinations, wherein the parameters included in the parameter value combination are the speed parameter, the cooling parameter, the cargo weight and the road condition parameter;

[0103] Based on the carbon emission prediction model of cold chain delivery vehicles, the carbon emissions corresponding to each parameter value combination are predicted, and the speed parameter value and the refrigeration parameter value corresponding to the parameter value combination with the minimum carbon emissions are output;

[0104] Among them, the carbon emission prediction model for cold chain delivery vehicles is obtained by training based on a combination of sample parameter values ​​with carbon emission labels.

[0105] In another aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, which is implemented when the computer program is executed by a processor to perform a method for optimizing parameters affecting carbon emissions of cold chain delivery vehicles provided by the above methods, the method comprising:

[0106] Obtain speed parameters, refrigeration parameters, cargo weight, and road condition parameters of cold chain delivery vehicles during driving;

[0107] Under the condition that the cargo weight and the road condition parameter remain unchanged, based on the change threshold of the speed parameter and the change threshold of the cooling parameter, the speed parameter value and the cooling parameter value are changed to generate a plurality of parameter value combinations, wherein the parameters included in the parameter value combination are the speed parameter, the cooling parameter, the cargo weight and the road condition parameter;

[0108] Based on the carbon emission prediction model of cold chain delivery vehicles, the carbon emissions corresponding to each parameter value combination are predicted, and the speed parameter value and the refrigeration parameter value corresponding to the parameter value combination with the minimum carbon emissions are output;

[0109] Among them, the carbon emission prediction model for cold chain delivery vehicles is obtained by training based on a combination of sample parameter values ​​with carbon emission labels.

[0110] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0111] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0112] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. 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 method for optimizing parameters affecting carbon emissions of cold chain delivery vehicles, characterized in that: include: Obtain speed parameters, refrigeration parameters, cargo weight, and road condition parameters of cold chain delivery vehicles during driving; Under the condition that the cargo weight and the road condition parameter remain unchanged, based on the change threshold of the speed parameter and the change threshold of the cooling parameter, the speed parameter value and the cooling parameter value are changed to generate a plurality of parameter value combinations, wherein the parameters included in the parameter value combination are the speed parameter, the cooling parameter, the cargo weight and the road condition parameter; Based on the carbon emission prediction model of cold chain delivery vehicles, the carbon emissions corresponding to each parameter value combination are predicted, and the speed parameter value and the refrigeration parameter value corresponding to the parameter value combination with the minimum carbon emissions are output; The carbon emission prediction model for cold chain delivery vehicles is obtained by training based on a combination of sample parameter values ​​with carbon emission labels; The cold chain delivery vehicle carbon emission prediction model is based on the iTransformer model by replacing the layer normalization equation with the depth normalization equation and replacing the multivariate attention mechanism with a sliding window attention mechanism that integrates global information.

2. The method for optimizing the parameters affecting carbon emissions of cold chain delivery vehicles according to claim 1, characterized in that: The road condition parameters of the cold chain delivery vehicle during driving are obtained in the following ways, including: Acquire a road condition image of the cold chain delivery vehicle during driving; Identify the road condition image using a road condition information recognition model to determine the number of vehicles, vehicle categories, and pixels of a vehicle prediction box in the road condition image; The pixel area ratio of the vehicle is determined based on the number of vehicles, the vehicle category and the pixels of the vehicle prediction frame, and the pixel area of ​​the vehicle is the road condition parameter.

3. The method for optimizing the parameters affecting carbon emissions of cold chain delivery vehicles according to claim 2, characterized in that: The pixel area ratio of the vehicle is determined based on the number of vehicles, the vehicle category and the pixels of the vehicle prediction frame, and is implemented by the following formula: Among them, i represents the vehicle category, S i represents the pixel area ratio of the i-th type of vehicle, n i is the number of vehicles of the i-th category in the road condition image, W ij is the horizontal pixel of the jth vehicle prediction box in the i-th vehicle, H ij is the vertical pixel of the prediction box of the jth vehicle in the i-th vehicle category, and Z is the pixel area of ​​the road condition image.

4. The method for optimizing the parameters affecting carbon emissions of cold chain delivery vehicles according to claim 2, characterized in that: The road condition information recognition model is based on the YOLOv8s model, introduces full-dimensional dynamic convolution in the convolution layer for deep feature extraction in the backbone network, and adds a progressive feature pyramid network between the convolution layer in the neck network and the convolution layer for deep feature extraction and between two convolution modules for deep feature extraction.

5. The method for optimizing the parameters affecting carbon emissions of cold chain delivery vehicles according to claim 1, characterized in that: The refrigeration parameters include refrigeration temperature, refrigeration humidity and refrigeration power.

6. A device for optimizing parameters affecting carbon emissions of cold chain delivery vehicles, characterized in that: include: An acquisition module is used to obtain speed parameters, refrigeration parameters, cargo weight and road condition parameters of the cold chain delivery vehicle during driving; A generating module, configured to change the speed parameter value and the cooling parameter value to generate a plurality of parameter value combinations based on the change threshold of the speed parameter and the change threshold of the cooling parameter while keeping the cargo weight and the road condition parameter unchanged, wherein the parameters included in the parameter value combination are the speed parameter, the cooling parameter, the cargo weight and the road condition parameter; The first processing module is used to predict the carbon emissions corresponding to each parameter value combination based on the carbon emission prediction model of the cold chain delivery vehicle, and output the speed parameter value and the refrigeration parameter value corresponding to the parameter value combination with the minimum carbon emissions; The carbon emission prediction model for cold chain delivery vehicles is obtained by training based on a combination of sample parameter values ​​with carbon emission labels; The cold chain delivery vehicle carbon emission prediction model is based on the iTransformer model by replacing the layer normalization equation with the depth normalization equation and replacing the multivariate attention mechanism with a sliding window attention mechanism that integrates global information.

7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, it implements the method for optimizing the parameters affecting the carbon emissions of the cold chain delivery vehicle as described in any one of claims 1 to 5.

8. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, a method for optimizing parameters affecting carbon emissions of a cold chain delivery vehicle as described in any one of claims 1 to 5 is implemented.

9. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, a method for optimizing parameters affecting carbon emissions of a cold chain delivery vehicle as described in any one of claims 1 to 5 is implemented.

Images