A method, apparatus, and device for torque distribution

Abstract

The application provides a torque distribution method, device and equipment applied to a hybrid vehicle, the method comprising: determining a plurality of output torques of a current engine speed at a preset engine efficiency from an engine speed, a value range of an output torque of the engine and a mapping relationship of engine efficiency; obtaining energy values of at least one candidate output torque meeting a preset condition in the plurality of output torques respectively by using a relationship model according to a current state of charge of a battery, a cooperative state variable and the at least one candidate output torque; selecting a target energy value smaller than a first preset threshold from the energy values of the plurality of output torques respectively, and determining an actual output torque of the current engine based on an output torque corresponding to the selected target energy value. Through the above method, an accurate distribution result can be obtained when the PMP algorithm is used to distribute the output torque of the engine.

Description

Technical Field

[0001] This application relates to the field of automotive control technology, specifically to a method, apparatus, and device for torque distribution. Background Technology

[0002] Hybrid electric vehicles are crucial for mitigating the global energy and environmental crisis and achieving leapfrog development in the automotive industry. Their multi-energy-source coupled drive configuration provides a foundation for further exploring the potential for energy conservation and emission reduction. A well-designed energy management strategy can improve fuel economy while ensuring vehicle performance.

[0003] The Pontryagin's Minimum Principle (PMP) algorithm's Equivalent Consumption Minimum Strategy (ECMS) can quickly solve the problem of optimal energy allocation in hybrid electric vehicles. However, the current PMP algorithm's Hamiltonian function construction is based solely on the research objective and cannot obtain accurate initial values ​​for the costate variables within the Hamiltonian function. This may result in inaccurate results when using the PMP algorithm to solve for optimal energy allocation. Summary of the Invention

[0004] This application provides a method, apparatus, and device for torque distribution, which can obtain accurate distribution results when using the PMP algorithm to distribute the output torque of an engine.

[0005] In a first aspect, embodiments of this application provide a torque distribution method, the method comprising:

[0006] Based on the range of engine speed, engine output torque, and the mapping relationship of engine efficiency, determine multiple output torques at the current engine speed under the preset engine efficiency.

[0007] Based on the battery's current state of charge, co-state variables, and at least one candidate output torque that meets preset conditions among the multiple output torques, the energy value of each of the at least one candidate output torque is obtained using the relational model; wherein, the co-state variables correspond to the vehicle's operating data; the relational model is used to describe the relationship between the engine's output torque, co-state variables, the battery's state of charge, and the energy value generated by the vehicle's operation, and the co-state variables are determined after transforming the relational model;

[0008] Select a target energy value that is less than a first preset threshold from the energy values ​​of the plurality of output torques, and determine the actual output torque of the current engine based on the output torque corresponding to the selected target energy value.

[0009] In the above embodiments, multiple output torques at a preset engine efficiency can be obtained using the current engine speed. Based on these multiple output torques, at least one candidate output torque that meets preset conditions can be obtained. Through a pre-built relational model, the energy value of each of the at least one candidate output torque can be obtained. A target energy value less than a first preset threshold is selected from these energy values, and the output torque corresponding to the target energy value is taken as the actual output torque of the current engine. This embodiment does not calculate the energy values ​​corresponding to all output torques, but instead selects a portion of the output torques that meet the conditions to calculate their energy values, saving unnecessary calculation steps and avoiding the influence of output torques that do not meet the preset conditions on the final result.

[0010] In one possible implementation, the relational model is constructed as follows:

[0011] The relationship model is obtained by summing the product of the engine fuel consumption function, the battery state of charge function and the costate variable, and the over-boundary penalty function.

[0012] The engine fuel consumption function is constructed based on the actual operating data of the engine, the battery state of charge function is constructed based on the actual operating data of the battery, and the over-boundary penalty function is constructed based on the preset output torque.

[0013] In the above embodiments, a superboundary penalty function based on a preset output torque is added to the relational model, enabling the relational model to identify whether the output torque is superboundary.

[0014] In one possible implementation, the at least one candidate output torque is determined in the following manner:

[0015] The multiple output torques are respectively input into the superboundary penalty function included in the relation model to obtain the function values ​​of the superboundary penalty function corresponding to the multiple output torques respectively;

[0016] Determine at least one candidate output torque from the plurality of output torques whose function value is less than a second preset threshold.

[0017] In the above embodiments, the superboundary penalty function included in the relational model can be used to determine whether multiple output torques meet preset conditions (whether they are within preset boundaries). If the function value of the superboundary penalty function corresponding to any output torque is not less than a second preset threshold, it indicates that the preset conditions are not met, that is, any output torque exceeds the boundary. Through the embodiments of this application, output torques within preset boundaries can be selected, thereby ensuring accurate results in subsequent calculations of energy values ​​and determination of actual output torques.

[0018] In one possible implementation, costate variables are determined in the following manner:

[0019] Based on the relationship between the co-state variables obtained by transforming the relational model and the vehicle operation data, the target co-state variable corresponding to the vehicle operation data at the current moment is determined.

[0020] In the above embodiments, the co-state variable is an unknown quantity, which can be determined by the relationship between the co-state variable obtained by transforming the relational model and the vehicle operation data.

[0021] In one possible implementation, the vehicle operating data includes the engine's output torque, and the current time is the time when the vehicle starts operating;

[0022] The step of determining the target covariate corresponding to the vehicle operation data at the current moment based on the relationship between the covariate obtained by transforming the relational model and the vehicle operation data includes:

[0023] Based on the correspondence between the costate variables obtained by transforming the pre-built relational model and the engine output torque, the target costate variables corresponding to each output torque at the current engine speed under the preset engine efficiency are determined.

[0024] In one possible implementation, determining the target costate variables corresponding to each output torque at the current engine speed under a preset engine efficiency, based on the correspondence between costate variables obtained by transforming a pre-built relational model and the engine's output torque, includes:

[0025] Based on the correspondence between the costate variables and the engine's output torque, candidate costate variables corresponding to each output torque are determined.

[0026] A binary search strategy is adopted to determine the target covariates corresponding to the output torques at the current engine speed under the preset engine efficiency from each candidate covariate.

[0027] In the above embodiments, if the current time is the time when the vehicle starts running, the correspondence between the co-state variables obtained by transforming the pre-built relational model and the engine's output torque can be used to determine multiple candidate co-state variables corresponding to the output torque. Then, a binary search strategy is used to select a target co-state variable from the candidate co-state variables. This application clarifies the process for determining the co-state variables at the time of vehicle start-up, avoiding blindly iterating to obtain the co-state variables and improving the efficiency of determining the co-state variables at the time of vehicle start-up.

[0028] In one possible implementation, the vehicle operation data includes the battery's state of charge, and the current time is any time other than the start time of vehicle operation.

[0029] The step of determining the target covariate corresponding to the vehicle operation data at the current moment based on the relationship between the covariate obtained by transforming the relational model and the vehicle operation data includes:

[0030] Based on the correspondence between the costate variables obtained by transforming the pre-built relational model and the state of charge of the battery, the target costate variable corresponding to the state of charge of the battery at the current moment is determined.

[0031] In the above embodiments, if the current time is any time other than the start time, the co-state variable at the current time can be obtained through the correspondence between the co-state variable and the battery's state of charge, without utilizing the method for determining the co-state variable at the start time. The method for calculating the co-state variable at other times provided in this application embodiment ensures the accuracy of the co-state variable at other times while improving the efficiency of torque distribution.

[0032] Secondly, embodiments of this application provide a torque distribution device, the device comprising:

[0033] Multiple output torque modules are defined to determine multiple output torques at the current engine speed and preset engine efficiency based on the mapping relationship between engine speed, the range of engine output torque values, and engine efficiency.

[0034] The energy value determination module is used to determine the energy value of each of the at least one candidate output torque based on the battery's current state of charge, co-state variables, and at least one candidate output torque that meets preset conditions among the multiple output torques, using the relational model; wherein, the co-state variables correspond to the vehicle's operating data; the relational model is used to describe the relationship between the engine's output torque, co-state variables, battery state of charge, and the energy value generated by vehicle driving, and the co-state variables are determined after transforming the relational model;

[0035] The actual output torque is determined by selecting a target energy value that is less than a first preset threshold from the energy values ​​of the plurality of output torques, and determining the actual output torque of the current engine based on the output torque corresponding to the selected target energy value.

[0036] Thirdly, embodiments of this application provide a torque distribution device, the device comprising:

[0037] At least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method as described in the first aspect above.

[0038] Fourthly, embodiments of this application provide a computer storage medium storing a computer program for causing a computer to perform the method described in the first aspect above. Attached Figure Description

[0039] Figure 1 This is a schematic flowchart illustrating a torque distribution method according to an exemplary embodiment of the present invention;

[0040] Figure 2 This is a schematic diagram of an equivalent model of a power battery according to an exemplary embodiment of the present invention;

[0041] Figure 3 This is a flowchart illustrating a method for determining costate variables at a starting time, as exemplified by an exemplary embodiment of the present invention.

[0042] Figure 4 This is a schematic diagram illustrating a specific process of a torque distribution method according to an exemplary embodiment of the present invention;

[0043] Figure 5 A schematic diagram of a torque distribution simulation system as an example of an exemplary embodiment of the present invention;

[0044] Figure 6 A schematic diagram of a torque distribution device according to an exemplary embodiment of the present invention;

[0045] Figure 7 This is a schematic diagram of a torque distribution device as an example of an exemplary embodiment of the present invention. Detailed Implementation

[0046] The technical solutions in the embodiments of this application will now be described clearly and in detail with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0047] The following is a description of the technical terms used in the embodiments of this application:

[0048] Pontryagin's Minimum Principle (PMP) is a theoretically based strategy for minimizing equivalent costs. The principle is that, given constraints on the control variable u, the objective function J is minimized, thereby solving for the optimal control variable.

[0049] In this embodiment, the J function, also known as the value function, is a cost function determined based on the research objective. Since this embodiment is based on the premise of battery state of charge (SOC) balance, the vehicle's energy consumption comes from the engine's fuel consumption, and the value function is the engine fuel consumption function. The research objective can be fuel economy or emissions compliance, etc., and is not specifically limited here. SOC balance means that the difference between the SOC at the start and end of vehicle operation is less than a preset value.

[0050] Hamiltonian (H) function: A function in the PMP algorithm, consisting of a value function, costate variables, and state constraints. It is a function of the control variable u. The number of costates is equal to the number of variables in the system model. In this embodiment, u represents the engine's output torque, and the state constraints of the H function are obtained based on the State of Charge (SOC).

[0051] The PMP algorithm is widely used in solving the problem of optimal energy allocation in hybrid power systems. However, the construction of the H function in the current PMP algorithm is based solely on the research objective and the costate variables at the initial time in the H function cannot be determined. This may result in inaccurate results when using the PMP algorithm to solve for optimal energy allocation.

[0052] To address the aforementioned problems, this application provides a torque distribution method based on the PMP algorithm, applicable to hybrid electric vehicles, such as... Figure 1 As shown, the method includes:

[0053] S101: Determine multiple output torques at the current engine speed and preset engine efficiency based on the range of engine speed, engine output torque, and the mapping relationship of engine efficiency.

[0054] Engine efficiency refers to the degree to which an engine effectively utilizes the thermal energy of fuel. For an engine, at the same engine speed, different engine efficiencies correspond to different output torques. For example, if the current engine speed is 1000 rpm, and the preset engine efficiency is 80% to 90%, the output torque at 1000 rpm will be between 200 N·m and 300 N·m. Then, multiple output torques are randomly selected from 200 N·m to 300 N·m. This application embodiment does not specifically limit the selection method; multiple output torques can be selected at fixed / non-fixed intervals. The precision of the selected values ​​can be accurate to the units place (e.g., 212 N·m) or to the tenths place (e.g., 204.5 N·m). Furthermore, this application embodiment does not specifically limit the number of selected output torques.

[0055] S102: Based on the battery's current state of charge, co-state variables, and at least one candidate output torque that meets preset conditions among the multiple output torques, the energy value of each of the at least one candidate output torque is obtained using the relational model.

[0056] This application embodiment determines the energy value of at least one candidate output torque based on the PMP algorithm. The main steps include constructing the H function (relational model), calculating the costate variables, and calculating the energy value. The specific implementation method is as follows:

[0057] 1. Construct the H function.

[0058] In this embodiment of the application, constructing the H function is equivalent to constructing a relational model. The relational model is used to describe the relationship between the engine's output torque, costate variables, battery state of charge, and the energy value generated by vehicle driving.

[0059] Unlike existing H functions, this application embodiment first constructs a super-boundary penalty function based on a preset output torque when constructing the H number. The super-boundary penalty function is shown in Formula 1:

[0060] D = A * | C current -C boderlimit | Formula 1

[0061] In Formula 1, D represents the superboundary penalty function, and A is a constant of a high order of magnitude, such as 10. 5 C current C represents any one of the multiple input torques in the above embodiments as the output torque. boderlimit This refers to the preset output torque, which is the limit value for output torque. For example, for a certain car engine, its maximum output torque cannot exceed 500 N·m. boderlimit The value is 500. Formula 1 means that if any output torque is less than the preset output torque, it indicates that the output torque is within the limits of the given model parameters, then let C... current and C boderlimit If the values ​​are equal, the function value D of the superboundary penalty function is 0. If any output torque is greater than the preset output torque, it indicates that any output torque exceeds the limit under the given model parameters. Then let C... current and C boderlimit If the values ​​are not equal, the function value D of the superboundary penalty function will be larger.

[0062] Using Formula 1 above, at least one candidate output torque whose function value of the out-of-boundary penalty function is less than the second preset threshold can be determined from the above multiple output torques, and then the subsequent step S103 is performed, avoiding unnecessary calculation steps, and also eliminating the influence of output torque exceeding the limit on the result. In addition to determining whether the output torque is out of range, the out-of-boundary penalty function provided in this application embodiment can also further determine whether the engine speed, motor speed, and motor torque are out of range.

[0063] In addition to the aforementioned super-boundary penalty function, the H function constructed in this application also includes an engine fuel consumption function, a battery state of charge function, and a costate variable. Specifically, the H function is obtained by summing the product of the engine fuel consumption function, the battery state of charge function, and the costate variable, as well as the super-boundary penalty function, as shown in Formula 2 below. The engine fuel consumption function is constructed based on the actual operating data of the engine, as shown in Formula 3, and the battery state of charge function is constructed based on the actual operating data of the battery, as shown in Formula 4.

[0064] H(u)=dJ+λ×dS+D Formula 2

[0065] In Formula 2, H(u) represents the H function, u represents the control variable, i.e., the output torque; J is the engine fuel consumption function; S is the battery state of charge function; λ is the co-state variable; D is the super-boundary penalty function, as in Formula 1, dJ represents the derivative of the engine fuel consumption function with respect to time, and dS represents the derivative of the battery state of charge function with respect to time.

[0066] J=∑m(n eng ,T eng ) Formula 3

[0067] In Formula 3, J represents the engine fuel consumption function, m is the total fuel consumption of the vehicle during driving, and n... eng T is the engine speed. eng This refers to the engine's output torque.

[0068]

[0069] In Formula 4, S represents the battery state-of-charge function, I b (t) represents the current of the power battery, since the power battery can be equivalent to, for example, Figure 2 The voltage-resistance model shown provides the formula for calculating the battery current, as shown in Formula 5. Therefore, based on the equivalent model, the formula for calculating the battery current, Q, can be obtained. b This represents the total capacity of the battery.

[0070]

[0071] In Formula 5, I bU(t) is the current of the power battery, U(t) is the open-circuit voltage of the power battery, P(t) is the output power of the power battery, and R(t) is the internal resistance of the power battery.

[0072] By constructing the H function using formulas 3 to 5 above, we obtain formula 6:

[0073]

[0074] In Formula 6, H(u) represents the H function, u represents the control variable, i.e., the output torque, m is the total fuel consumption of the vehicle during driving, and n... eng T is the engine speed. eng Let λ be the engine's output torque, λ be the costate variable, U(t) be the open-circuit voltage of the power battery, P(t) be the power battery's output power, R(t) be the power battery's internal resistance, and Q be the output torque of the engine. b The total capacity of the battery is A, where A is a constant of a high order of magnitude, such as 10. 5 C current For any one of the multiple input torques, C boderlimit This is the preset output torque, which is the limit value of the output torque.

[0075] 2. Calculate the costate variables.

[0076] Costate variables are parameters used to construct the H-function and correspond to vehicle operating data. For example, costate variables correspond to engine output torque and battery state of charge. As shown in the above implementation method for constructing the H-function, the pre-constructed relational model includes costate variables. To achieve torque distribution during vehicle operation, the costate variables need to be determined first. Since costate variables correspond to vehicle operating data, they can be determined based on the relationship between the costate variables obtained by transforming the relational model and the vehicle operating data. The specific implementation method is as follows:

[0077] In this application embodiment, the methods for determining costate variables are divided into the following two types:

[0078] (1) The current time is the time when the vehicle starts running.

[0079] The starting time of the vehicle is the initial time during the vehicle's operation. For example, if the vehicle is running between 9:00 and 10:00, the starting time of the vehicle is 9:00.

[0080] For torque distribution using the PMP algorithm, solving for the costate variables at the initial running moment of the vehicle is crucial. The specific process is as follows: Figure 3 As shown,

[0081] S301: Based on S101 above, obtain multiple output torques;

[0082] S302: Simplify the relational model by simplifying Formula 2 above to obtain H1(u)=dJ+λ×dS, where H1(u) represents the simplified H function, dJ represents the derivative of the engine fuel consumption function with respect to time, dS represents the derivative of the battery state of charge function with respect to time, and λ is a costate variable.

[0083] S303: Further transform the simplified relational model to obtain the correspondence between the costate variables and the engine's output torque. Specifically, differentiate H1(u) with respect to u, and then set the derivative function to 0, that is, set d(dJ) / du+λ×d(dS) / du=0, to obtain the correspondence between the costate variables and the engine's output torque:

[0084]

[0085] Where d(dJ) / du represents the double derivative of the engine fuel consumption function with respect to time and output torque, d(dS) / du represents the double derivative of the battery state-of-charge function with respect to time and output torque, u represents the output torque, and λ is a costate variable;

[0086] S304: Based on the correspondence between costate variables and engine output torque, determine the candidate costate variables corresponding to each output torque. For example, if there are multiple output torques of 100 N.m, 105 N.m, 110 N.m, 115 N.m, and 120 N.m, use Formula 7 to calculate the costate variable corresponding to 100 N.m as a, the costate variable corresponding to 105 N.m as b, the costate variable corresponding to 110 N.m as c, the costate variable corresponding to 100 N.m as d, and the costate variable corresponding to 100 N.m as e. Use a, b, c, d, and e as candidate costate variables.

[0087] S305: The binary search strategy is adopted to determine the target covariates corresponding to the output torques of the engine speed at the current moment under the preset engine efficiency from the candidate covariates.

[0088] Binary search is a highly efficient algorithm, especially when dealing with large amounts of data. Its time complexity is log(n), where n is the total amount of data. The main idea of ​​binary search is to repeatedly fold all the data in half, removing half of the data in each search, until finally all results that do not meet the criteria are removed, leaving only one result that meets the criteria.

[0089] The conditions set in this embodiment are as follows: if the difference between the current state of charge (SOC) of the battery and the SOC obtained based on an arbitrary costate variable at the next moment is less than a preset value, then the arbitrary costate variable is determined as the target costate variable. The SOC of the battery at the next moment corresponding to the arbitrary costate variable can be obtained using formulas 8 and 9.

[0090] λ t+1 =λ t +Δt×λ' t+1 Formula 8

[0091]

[0092] In Equations 8 and 9, H represents the H function, and λ t+1 Let λ be the costate variable at the next time step. t Let λ' be the costate variable at the current time. t+1 This represents the differentiation of the costate variable with respect to the next moment, where SOC(t+1) is the battery state of charge at the next moment, and Δt is the time interval between the current moment and the previous moment.

[0093] First, based on the relationship between the covariate at the current moment and the covariate at the next moment (Formula 8), the derivative of the covariate at the next moment is obtained. Then, based on the derivative of the covariate at the next moment, the state of charge of the battery at the next moment can be predicted (Formula 9). If the difference between the state of charge of the battery at the current moment and the state of charge of the battery at the next moment is less than a preset value, then the arbitrary covariate can be determined as the target covariate.

[0094] Based on the example of S304 above, the candidate costate variables are a, b, c, d, and e. In this embodiment, a, b, c, d, and e are divided into two groups to obtain array 1 [a, b, c] and array 2 [d, e]. First, it is determined whether d and e in array 2 meet the conditions according to formulas 8 and 9 above. If they do not meet the conditions, the target costate variable is determined to be in array 1. Then, array 1 is further divided into groups to obtain array 1.1 [a, b] and array 1.2 [c]. It is then determined whether the elements in array 1.2 meet the conditions. If d in array 1 meets the conditions, then d can be used as the target costate variable, without needing to traverse the elements in array 1 again.

[0095] Based on the characteristic of "multiple folding of data" in the binary search strategy, this application embodiment also sets "iteration count" and "iteration precision". When the iteration count condition or the iteration precision condition is met, the folding stops.

[0096] For example, if the "iteration count" is set to 100, the 100th iteration determines the target covariate from the covariates at positions 6 to 8. It checks whether the 6th, 7th, and 8th covariates meet the conditions set in the above embodiment. The covariates that meet the conditions are taken as the target covariates. If there are multiple covariates that meet the conditions, the covariate with the smaller difference is taken as the target covariate.

[0097] For example, if the "iteration precision" is set to be accurate to the units place, the target covariate is determined from the covariates at positions 6 to 7 in the current iteration. Since (6+7) / 2 = 6.5, the precision of the next iteration is determined to be the tenths place, which is less than the preset precision. At this time, the iteration stops, and the covariate that satisfies the conditions in the above embodiment is selected from the 6th and 7th covariates as the target covariate. If there are multiple covariates that satisfy the conditions, the covariate with the smaller difference is selected as the target covariate.

[0098] (2) The current time is any time other than the time when the vehicle starts running during operation.

[0099] For example, if a vehicle operates between 9:00 and 10:00, the vehicle starts operating at 9:00, and other times are times other than 9:00, such as 9:01, 9:02, 9:03...10:00.

[0100] After obtaining the costate variables at the initial running time of the vehicle, in order to simplify subsequent calculations, the costate variables at the initial time can be updated according to the correspondence between the costate variables obtained by transforming the pre-built relational model and the state of charge of the battery, so as to obtain the costate variables at the next time.

[0101] The derivation process of the correspondence between the costate variable and the state of charge of the battery is shown in Formulas 10 to 12 below:

[0102]

[0103] λ t =λ t-1 +Δt×λ' t Formula 11

[0104]

[0105] Where SOC(t) is the current state of charge of the battery, Δt is the time interval between the current time and the previous time, and λ t Let λ be the costate variable at the current time. t-1 Let λ' be the costate variable of the previous time step. t Indicates the relationship with λ tPerform a differentiation operation. Since the differentiation of the costate variable at the current moment is related to both the battery's state of charge and the costate variable at the current moment, the correspondence between the costate variable and the battery's state of charge can be obtained through the differentiation of the costate variable. As shown in Equation 12, the costate variable at the current moment can be obtained from the battery's state of charge at the current moment and the costate variable at the previous moment.

[0106] For example, the costate variable 'a' in the previous time step has a current state of 0.5 (SOC). Given a = 2, and the time interval between the current moment and the previous moment is 1 minute, substituting a into formula 12, we can obtain the costate variable at the current moment, such as a + 1 × 2 = f.

[0107] 3. Calculate the energy value of at least one candidate output torque.

[0108] After determining at least one candidate output torque and costate variable through the implementation method in S102 above, the energy value of each candidate output torque, that is, the function value of the H function, is obtained by using the relational model.

[0109] For example, if any candidate output torque is 200 N·m and the costate variable is a, using Formula 6, the energy value is 100 Wh / kg, that is, the energy value corresponding to the candidate output torque of 200 N·m is 100 Wh / kg.

[0110] After obtaining multiple energy values, it is necessary to select a target energy value from the multiple energy values. The specific implementation method is as follows: S103.

[0111] In addition, the electrical energy generated by the battery is converted into mechanical energy to drive the vehicle's motor. The motor and engine are used to drive the rotation of the car's wheels. Therefore, when calculating the energy value, it is not simply a matter of inputting the candidate output torque and co-state variables into Formula 6. Instead, it is necessary to consider the power transmission between various vehicle components, such as the energy transfer between the battery, motor, and wheels.

[0112] For the energy management strategy of hybrid vehicles, the lateral and vertical motion of the vehicle can be ignored, and only the longitudinal motion can be studied. Furthermore, the slippage between the wheels and the ground can be ignored. The torque required at the wheel end at different times under the known parameters such as vehicle speed and wheel radius can be obtained as Equation 13.

[0113]

[0114] In Formula 13, T whl For the torque required at the wheel end, F r The resistance to movement can be determined from the vehicle's operating data obtained during the experiment, in meters (m). veh For the overall vehicle weight, v'veh For the acceleration of the whole vehicle, r whl J is the radius of the entire wheel end. whl Let w' be the moment of inertia of the entire wheel end. whl The wheel end angle acceleration, For a weighted system, J k Let w' be the moment of inertia of a certain driving component, which includes an engine and a motor, and k be the number of driving components, for example, 2. k It represents the rotational angular velocity of a certain driving component.

[0115] The energy management process only studies the engine's fuel economy. While meeting the model accuracy requirements, the engine's high-frequency dynamic characteristics are ignored to improve simulation speed and reduce simulation calculation costs. A quasi-static model of the engine is established based on engine bench test data, as shown in Equation 14.

[0116] b eng =C fuel (n eng ,T eng ) Formula 14

[0117] In formula 14, b eng For fuel consumption rate, C fuel Let n be the engine fuel consumption function. eng T is the engine speed. eng This refers to the engine's output torque.

[0118] The actual output torque of the motor is related to the characteristics of both the motor and the power battery pack. The maximum driving torque and maximum generating torque of the motor are related to its speed, while the motor efficiency is related to its speed and torque. The calculation formula is Equation 15.

[0119] η m =f(n) m ,T m ) Formula 15

[0120] In Formula 15, η m For motor efficiency, n m T is the motor speed. m This represents the motor torque.

[0121] Formulas 13 to 15 and formulas 10 to 12 can be used to analyze how energy is transferred in various automotive components, thereby obtaining accurate energy values. The specific methods for solving energy values ​​are common knowledge in this field and will not be elaborated here.

[0122] S103: Select a target energy value that is less than a first preset threshold from the energy values ​​of the plurality of output torques, and determine the actual output torque of the current engine based on the output torque corresponding to the selected target energy value.

[0123] When determining the target energy value, a preset threshold can be set, and an energy value can be selected from the energy values ​​less than the preset threshold as the target energy value, or the smallest energy value can be selected as the target energy value.

[0124] After determining the target energy value, the candidate output torque corresponding to the target energy value can be further determined, and the candidate output torque can be used as the actual output torque of the current engine.

[0125] For example, the energy value corresponding to a candidate torque of 200 N·m is 100 Wh / kg, the energy value corresponding to 205 N·m is 90 Wh / kg, the energy value corresponding to 210 N·m is 110 Wh / kg, and the energy value corresponding to 215 N·m is 120 Wh / kg. The preset threshold is 95 Wh / kg, and the energy value less than 95 Wh / kg is 90 Wh / kg, or the smallest energy value among them is selected as 90 Wh / kg. Then, based on 90 Wh / kg, the candidate output torque is determined to be 205 N·m, and 205 N·m is taken as the actual output torque of the current engine.

[0126] The following is based on Figure 4 This application provides a detailed description of a torque distribution method according to an embodiment.

[0127] S401: Obtain multiple output torques at the current engine speed under a preset engine efficiency;

[0128] S402: Is the current time the time when the vehicle starts running? If yes, execute S403; otherwise, execute S408. The time when the vehicle starts running is the initial time during the vehicle's operation. For example, if the vehicle runs between 9:00 and 10:00, the time when the vehicle starts running is 9:00.

[0129] S403: Determine the costate variables at the current moment based on the correspondence between the costate variables obtained by transforming the pre-built relational model and the engine output torque;

[0130] S404: Select at least one candidate output torque from multiple output torques whose function value of the superboundary penalty function is less than a second preset threshold;

[0131] S405: Based on the battery's current state of charge, co-state variables, and at least one candidate output torque, the energy values ​​of each of the at least one candidate output torque are obtained using the aforementioned relationship model.

[0132] S406: Select a target energy value that is less than a first preset threshold from the energy values ​​of multiple output torques;

[0133] S407: Determine the actual output torque of the current engine based on the output torque corresponding to the target energy value, and end this process.

[0134] S408: Based on the correspondence between the costate variables obtained by transforming the pre-built relational model and the state of charge of the battery, determine the costate variables at the current moment, and execute S404 to S407.

[0135] This application provides a torque distribution method by adding a super-boundary penalty function to the H function of the PMP algorithm, which allows for reasonable handling of super-boundary situations during torque distribution, thereby making the results more accurate. This application also clarifies the calculation process of costate variables in the PMP algorithm, avoiding blind iterative solutions to costate variables, accelerating the search for costate variables at the initial moment, and improving the running efficiency of the PMP algorithm.

[0136] Based on the same inventive concept, embodiments of this application also provide a torque distribution system for simulating the torque distribution process, such as... Figure 5 As shown, it includes an information input module, a physical model module, and a control strategy model module.

[0137] The information input module inputs vehicle operating data, such as vehicle speed and engine speed, into the control strategy model module and the physical model module. The physical model module realizes the operation of the whole vehicle in the simulation environment. The control strategy model module performs energy management control based on the information provided by the information input module and the physical model module, and outputs relevant control signals to the physical model module. The physical model module and the control strategy model module form a closed-loop feedback control.

[0138] Among them, the battery state of charge planning module in the control strategy model module is used to realize the trajectory planning of battery SOC and solve the co-state variables, as shown in Formula 12. The co-state variables are then passed to the torque distribution module, whereby the PMP algorithm of this module performs torque distribution in real time and passes the results to the physical model module for execution.

[0139] The gear planning module and clutch planning module plan the gear and clutch based on the information fed back by the information input module, such as the current gear, current clutch, current engine speed, and current engine torque. Gear planning and clutch planning are existing technologies and will not be described in detail here.

[0140] Based on the same inventive concept, embodiments of this application also provide a torque distribution device, such as... Figure 6 As shown, the device includes:

[0141] Multiple output torque modules 601 are used to determine multiple output torques at the current engine speed and preset engine efficiency from the mapping relationship between engine speed, the range of engine output torque values ​​and engine efficiency.

[0142] The energy value determination module 602 is used to determine the energy value of each of the at least one candidate output torque based on the current state of charge of the battery, the co-state variable, and at least one candidate output torque that meets preset conditions among the plurality of output torques, using the relational model; wherein, the co-state variable has a corresponding relationship with the vehicle's operating data; the relational model is used to describe the relationship between the engine's output torque, the co-state variable, the battery's state of charge, and the energy value generated by the vehicle's driving, and the co-state variable is determined after transforming the relational model;

[0143] The actual output torque determination module 603 is used to select a target energy value that is less than a first preset threshold from the energy values ​​of the plurality of output torques, and determine the actual output torque of the current engine based on the output torque corresponding to the selected target energy value.

[0144] In one possible implementation, the device further includes a construction module for constructing the relational model in the following manner:

[0145] The relationship model is obtained by summing the product of the engine fuel consumption function, the battery state of charge function and the costate variable, and the over-boundary penalty function.

[0146] The engine fuel consumption function is constructed based on the actual operating data of the engine, the battery state of charge function is constructed based on the actual operating data of the battery, and the over-boundary penalty function is constructed based on the preset output torque.

[0147] In one possible implementation, the energy value determination module 602 is used to determine the at least one candidate output torque in the following manner:

[0148] The multiple output torques are respectively input into the superboundary penalty function included in the relation model to obtain the function values ​​of the superboundary penalty function corresponding to the multiple output torques respectively;

[0149] Determine at least one candidate output torque from the plurality of output torques whose function value is less than a second preset threshold.

[0150] In one possible implementation, the energy value determination module 602 is used to determine the costate variable in the following manner:

[0151] Based on the relationship between the co-state variables obtained by transforming the relational model and the vehicle operation data, the target co-state variable corresponding to the vehicle operation data at the current moment is determined.

[0152] In one possible implementation, the vehicle operating data includes the engine's output torque, and the current time is the time when the vehicle starts operating;

[0153] The energy value determination module 602 is used to determine the target co-state variables corresponding to each output torque at the current engine speed under the preset engine efficiency, based on the correspondence between the co-state variables obtained by transforming the pre-built relational model and the engine output torque.

[0154] In one possible implementation, the energy value determination module 602 is used to determine candidate costate variables corresponding to each output torque based on the correspondence between the costate variables and the engine's output torque.

[0155] A binary search strategy is adopted to determine the target covariates corresponding to the output torques at the current engine speed under the preset engine efficiency from each candidate covariate.

[0156] In one possible implementation, vehicle operation data includes the battery's state of charge, and the current time is any time other than the start time of vehicle operation.

[0157] The energy value determination module 602 is used to determine the target costate variable corresponding to the current state of charge of the battery based on the correspondence between the costate variable obtained by transforming the pre-built relational model and the state of charge of the battery.

[0158] Based on the same inventive concept, embodiments of this application also provide a torque distribution device, the device comprising:

[0159] At least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform a torque distribution method.

[0160] like Figure 7 As shown, the device includes a processor 701, a memory 702, and a communication interface 703; a bus 704. The processor 701, memory 702, and communication interface 703 are interconnected via the bus 704.

[0161] The processor 701 is configured to read and execute instructions from the memory 702, so that the at least one processor can perform the torque distribution method provided in the above embodiments.

[0162] The memory 702 is used to store various instructions and programs for the torque distribution method provided in the above embodiments.

[0163] The communication interface 703 is used for data interaction between the transient smoke sensor and the electronic control unit.

[0164] The 704 bus can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of representation, Figure 5 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0165] The processor 701 can be a central processing unit (CPU), a network processor (NP), a graphics processing unit (GPU), or any combination of CPU, NP, and GPU. It can also be a hardware chip. The aforementioned hardware chip can be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The aforementioned PLD can be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL), or any combination thereof.

[0166] Based on the same inventive concept, this application also provides a vehicle, the vehicle comprising:

[0167] Sensors are used to collect engine speed data in real time and send it to the electronic control unit;

[0168] The electronic control unit receives the engine speed from the sensor and executes the torque distribution method described above based on the engine speed.

[0169] In addition, this application also provides a computer-readable storage medium storing a computer program for causing a computer to perform a torque distribution method according to the above embodiments.

[0170] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0171] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0172] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0173] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A method for torque distribution, characterized in that, Applied to hybrid vehicles, the method includes: Based on the range of engine speed, engine output torque, and the mapping relationship of engine efficiency, determine multiple output torques at the current engine speed under the preset engine efficiency. The relationship model is obtained by summing the product of the engine fuel consumption function, the battery state of charge function and the costate variable, and the over-boundary penalty function. The engine fuel consumption function is constructed based on the actual operating data of the engine, the battery state of charge function is constructed based on the actual operating data of the battery, and the over-boundary penalty function is constructed based on the preset output torque. Based on the battery's current state of charge, co-state variables, and at least one candidate output torque that meets preset conditions among the multiple output torques, the energy value of each of the at least one candidate output torque is obtained using the relational model; wherein, the co-state variables correspond to the vehicle's operating data; the relational model is used to describe the relationship between the engine's output torque, co-state variables, the battery's state of charge, and the energy value generated by the vehicle's operation, and the co-state variables are determined after transforming the relational model; Select a target energy value that is less than a first preset threshold from the energy values ​​of each of the at least one candidate output torque, and determine the actual output torque of the current engine based on the output torque corresponding to the selected target energy value.

2. The method according to claim 1, characterized in that, The at least one candidate output torque is determined in the following manner: The multiple output torques are respectively input into the superboundary penalty function included in the relation model to obtain the function values ​​of the superboundary penalty function corresponding to the multiple output torques respectively; Determine at least one candidate output torque from the plurality of output torques whose function value is less than a second preset threshold.

3. The method according to claim 1, characterized in that, Costate variables are determined in the following manner: Based on the relationship between the co-state variables obtained by transforming the relational model and the vehicle operation data, the target co-state variable corresponding to the vehicle operation data at the current moment is determined.

4. The method according to claim 3, characterized in that, The vehicle operating data includes the engine's output torque, and the current time is the moment when the vehicle starts running. The step of determining the target covariate corresponding to the vehicle operation data at the current moment based on the relationship between the covariate obtained by transforming the relational model and the vehicle operation data includes: Based on the correspondence between the costate variables obtained by transforming the pre-built relational model and the engine output torque, the target costate variables corresponding to each output torque at the current engine speed under the preset engine efficiency are determined.

5. The method according to claim 4, characterized in that, The method for determining the target costate variables corresponding to each output torque at the current engine speed under a preset engine efficiency, based on the correspondence between the costate variables obtained by transforming a pre-built relational model and the engine's output torque, includes: Based on the correspondence between the costate variables and the engine's output torque, candidate costate variables corresponding to each output torque are determined. A binary search strategy is adopted to determine the target covariates corresponding to the output torques at the current engine speed under the preset engine efficiency from each candidate covariate.

6. The method according to claim 3, characterized in that, The vehicle operation data includes the battery's state of charge, and the current time is any time other than the start time of vehicle operation. The step of determining the target covariate corresponding to the vehicle operation data at the current moment based on the relationship between the covariate obtained by transforming the relational model and the vehicle operation data includes: Based on the correspondence between the costate variables obtained by transforming the pre-built relational model and the state of charge of the battery, the target costate variable corresponding to the state of charge of the battery at the current moment is determined.

7. A torque distribution device, characterized in that, The device includes: Multiple output torque modules are defined to determine multiple output torques at the current engine speed and preset engine efficiency based on the mapping relationship between engine speed, the range of engine output torque values, and engine efficiency. The energy value determination module is used to sum the product of the engine fuel consumption function, the battery state of charge function, and the co-state variable, as well as the over-boundary penalty function, to obtain a relational model. Based on the current state of charge of the battery, the co-state variable, and at least one candidate output torque that meets preset conditions among the multiple output torques, the energy value of each of the at least one candidate output torque is obtained using the relational model. The co-state variable corresponds to the vehicle's operating data. The engine fuel consumption function is constructed based on the actual operating data of the engine, the battery state of charge function is constructed based on the actual operating data of the battery, and the over-boundary penalty function is constructed based on the preset output torque. The relational model describes the relationship between the engine's output torque, the co-state variable, the battery state of charge, and the energy value generated by vehicle operation. The co-state variable is determined after transforming the relational model. The actual output torque is determined by selecting a target energy value less than a first preset threshold from the energy values ​​of the at least one candidate output torque, and determining the actual output torque of the current engine based on the output torque corresponding to the selected target energy value.

8. A torque distribution device, characterized in that, The device includes: At least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method as described in any one of claims 1-6.

9. A computer storage medium, characterized in that, The computer storage medium stores a computer program that enables the computer to perform the method as described in any one of claims 1-6.

Images