A low temperature cold start control method based on high frequency injection

Abstract

The application relates to the technical field of mechatronic systems, and particularly discloses a low-temperature cold start control method based on high-frequency injection, which comprises the following steps: analyzing the motor state, observing and feedforward compensating the disturbance caused by the load torque and the viscous coefficient to the motor state through a sliding mode observer algorithm; obtaining a state equation according to the motor state, constructing an extended sliding mode observer equation, deriving an observer error equation, and observing the load torque; analyzing the obtained load torque, and reflecting the oil viscosity through the motor load torque; injecting high-frequency current under the low-temperature state according to the oil viscosity result, controlling the injection frequency, enhancing the skin effect, controlling the current amplitude, utilizing the copper loss and the high-frequency magnetic field eddy current effect to assist heat generation, gradually reducing the injection voltage amplitude according to the final oil viscosity result, the electromagnetic heat generation viscosity and the load reduction, and switching to normal id=0 control after reaching the normal start condition.

Description

Technical Field

[0001] This application relates to the field of electromechanical-hydraulic systems technology, and specifically discloses a low-temperature cold start control method based on high-frequency injection. Background Technology

[0002] With the development of industrial automation and intelligent manufacturing, electromechanical-hydraulic systems have been widely used in aerospace, rail transportation, intelligent equipment and engineering machinery.

[0003] Instruction manual attached Figure 1 This paper illustrates the structure of an electromechanical-hydraulic system for off-road vehicles. Such systems typically consist of a motor, gear pump, valve block, and a closed-loop hydraulic medium, exhibiting a compact structure and high power density. However, they present fundamental challenges in low-temperature environments. When temperatures drop below -20°C, the viscosity of the hydraulic oil increases dramatically, hindering circulation, while the starting load torque of the pump increases by 5 to 10 times. Traditional motor control strategies exhibit serious deficiencies under these conditions. Open-loop V / F control fails to provide sufficient starting torque, leading to motor stall. The lack of auxiliary heating makes it difficult to reduce oil viscosity, significantly reducing system reliability, impacting system efficiency, and even causing malfunctions.

[0004] Instruction manual attached Figure 2 An electromechanical-hydraulic system is illustrated, with a submersible motor (ESP), gear pump, solenoid valve block, and closed-loop hydraulic oil as its core components, forming an energy transfer chain that couples "electrical energy - mechanical energy - hydraulic energy" across multiple physical domains. As the carrier of heat dissipation in the system, the thermal properties of the hydraulic oil directly affect the temperature rise and operating efficiency of the entire system. At low temperatures, the hydraulic oil solidifies, resulting in significant viscous resistance.

[0005] Currently, there are two main types of methods for low-temperature cold start of electromechanical-hydraulic systems: one is to study motor heating, using the inverter to output low-frequency AC current or DC bias current to heat the motor windings through copper losses; the other is to preheat the oil, embedding heating elements in the oil tank and pipelines, or integrating induction heating coils into the pump body. Engine waste heat is used to preheat the oil through a heat exchanger.

[0006] However, relying solely on the low-frequency AC current output of the motor inverter for heating via copper losses can lead to false triggering of overcurrent protection, resulting in startup failure. Adding heating elements and integrating induction heating into the pump body can cause localized rapid heating of the electric heating element, while also increasing costs. Furthermore, the actual system requires coordinated electromechanical-hydraulic control to predict low-temperature startup performance and optimize control parameters.

[0007] The following problems exist:

[0008] 1. The motor heating method may falsely trigger overcurrent protection, leading to startup failure. This method utilizes a low-frequency AC or DC bias current output from an inverter to heat the motor windings through copper losses. The preheating power needs to be dynamically adjusted based on real-time temperature. However, the motor's thermal parameters, such as heat capacity and thermal conductivity, change non-linearly with temperature, resulting in insufficient model accuracy. Furthermore, the low-frequency heating current generates high-order harmonics, interfering with onboard sensors and communication.

[0009] 2. Oil preheating is slow and time-consuming. Heat transfer from the oil tank to the actuator valve takes longer, resulting in a delayed system response. Exposed piping leads to rapid heat dissipation and low heating efficiency. While overall preheating requires heating all the oil, only the pump inlet and critical valve blocks require priority heating.

[0010] 3. Both lack effective coordination and fail to consider the overall electromechanical-hydraulic system. Most current studies only use single heating, with the heating system and hydraulic system separate, without overall optimization of total losses. At the same time, the passive response mode relies on current temperature feedback and cannot predict environmental changes or adjust strategies based on downtime.

[0011] 4. It is difficult to fully consider the limitations of the resistance material characteristics under extreme operating conditions, such as a sudden drop in the heating power of conventional PTC and a near loss of oil fluidity. Motor heating, due to the starting load torque exceeding viscous resistance, causes a sudden change in load torque, leading to a mismatch in preheating parameters and failure to heat according to the set conditions.

[0012] Therefore, there is an urgent need for a low-temperature cold start method to achieve electromechanical-hydraulic coordinated control, improving the reliability of the electromechanical-hydraulic system during low-temperature cold starts while maintaining a fast response speed. Especially in scenarios involving extremely low temperatures, high oil viscosity, and significant motor stall, this method can improve the reliability and response speed of the electromechanical-hydraulic system during cold starts. In view of this, the inventors propose a low-temperature cold start control method based on high-frequency injection. Summary of the Invention

[0013] The main objective of this invention is to overcome the problems of insufficient reliability, slow start-up speed, and inability to effectively coordinate and consider the overall electromechanical-hydraulic system conditions in the prior art during low-temperature cold starts. It provides a low-temperature cold start control method based on high-frequency injection to achieve rapid, reliable, and coordinated start-up of electromechanical-hydraulic systems under low-temperature conditions.

[0014] To achieve the above objectives, the present invention provides the following basic solution:

[0015] A low-temperature cold start control method based on high-frequency injection includes the following steps:

[0016] S101: Analyze the motor state and use the sliding diaphragm observer algorithm to observe and feedforward the disturbances caused by the motor state to the load torque and viscosity coefficient;

[0017] S102: Based on the motor state, obtain the state equation, construct the extended sliding diaphragm observer equation, derive the observer error equation, and observe the load torque;

[0018] S103: Analyze the obtained load torque. The motor load torque reflects the oil viscosity.

[0019] S104: Based on the oil viscosity results, high-frequency current injection is performed at low temperature to control the injection frequency, enhance the skin effect, and control the amplitude to utilize copper loss and high-frequency magnetic field eddy current effect to assist in heat generation.

[0020] S105: After high-frequency current injection, the motor generates copper losses. At low temperatures, the winding resistance decreases. High-frequency current causes the skin effect, which increases resistance and copper losses.

[0021] S106: After high-frequency current injection, the motor generates iron loss. The maximum iron loss point is found by continuously increasing the frequency of the high-frequency current.

[0022] S107: Motor copper loss causes winding temperature to rise. After high-frequency current injection, the motor generates iron loss, which causes the iron core to heat up. The heat is conducted to the casing and the oil temperature is increased by heating the oil pump casing.

[0023] S108: Based on the final oil viscosity result, the electromagnetic heating viscosity decreases and the load is reduced, so the injection voltage amplitude is gradually reduced. After the normal start-up conditions are met, the control switches to normal id=0.

[0024] Furthermore, the motor status includes the motor mechanical angle, speed, and load torque. Based on the motor status obtained by control with id=0, the given speed is compared with the feedback speed and output to the speed regulator. Given id=0, the current controller controls the d and q axis voltage inputs. The high-frequency pulsating voltage is superimposed on the d axis voltage, causing the d axis current to pulsate. The input d and q axis voltages undergo dq\αβ coordinate transformation and then space vector pulse width modulation to control the three-phase inverter. The DC power supply is output to the motor through the three-phase inverter to control the motor. The position sensor obtains the motor mechanical angle and speed by collecting position signals, the current sensor collects the current and feeds it back to the control loop, and the torque observer receives the current signal and speed signal to observe the load torque.

[0025] Furthermore, in step S102, the state equation is obtained, as follows:

[0026] Motor state equations:

[0027] In the formula For the estimated rotor mechanical angular velocity, Let J be the estimated load torque, J be the moment of inertia, and T be the moment of inertia. eFor electromagnetic rotation, B is the momentary viscous friction coefficient, and λ1 is the sliding mode observer gain. For the rotational speed observation error, sat() is a saturation function used to suppress sliding mode chattering, where:

[0028] The motor controller's sampling frequency is higher than the load torque change time relative to the load torque change time, assuming the load torque is a constant value within the control cycle, i.e.:

[0029] Furthermore, in step S102, combining the motion equations in the mathematical model of the permanent magnet synchronous motor, the mechanical angular velocity ω of the motor is... m and load torque T L The electromagnetic torque T is a state variable. e The input is the mechanical angular velocity ω, and the output is also the mechanical angular velocity. m The state equations are as follows:

[0030]

[0031] Based on the state equation, the extended synovial observer equation is constructed as follows:

[0032]

[0033] In the formula This is an estimate of the mechanical angular velocity; Load torque estimate; l is the feedback gain; U is the sliding mode control law. Subtracting the above two equations gives the observer error equation:

[0034] The observer error equation is as follows:

[0035] In the formula This represents the speed estimation error; To account for the load torque observation error, let the sliding mode switching surface s = e1, then U = -λsgn(s) - γs, which is obtained from the sliding mode attainability condition. We can obtain:

[0036]

[0037] In the formula, γ≥0, therefore we have Pick β≥1;

[0038] Combining the observer error equation, we obtain

[0039] Where c e Since it is a constant, only when l < 0 can e2 be guaranteed to approach zero, thus obtaining the load torque.

[0040] Further, in step S104, a historical reference database is established. This database contains known viscosity state samples. Given a specific temperature, rotational speed, and the torque value measured under that operating condition, the corresponding viscosity estimate can be queried or calculated. Simultaneously, real-time torque signals at stable rotational speeds are acquired. These signals are then processed to obtain the actual measured torque value at that rotational speed. Within the established historical database, based on the current rotational speed and load torque, the expected reference torque value at that temperature and rotational speed is searched or interpolated. The current oil viscosity estimate is then calculated and output using this reference torque value.

[0041] Furthermore, in step S014, the high-frequency voltage equation of PMSM in the dq coordinate system is first required; secondly, the expression of the high-frequency response current is obtained based on the high-frequency voltage equation of PMSM; then, rotating voltage injection is used to represent the injection voltage in matrix form and combine it with the expression of the high-frequency response current to obtain the high-frequency response current of HF-RSVI and the current response demodulation.

[0042] Furthermore, the high-frequency response current of the HF-RSVI is obtained as follows:

[0043] First: The high-frequency voltage equation of PMSM can be expressed as:

[0044]

[0045] In the formula, L0=(L d +L q ) / 2、L1=(L d -L q ) / 2, L d L q These are the d-axis and q-axis stator inductances, respectively.

[0046] Then: Inverse and integral the high-frequency voltage equation of PMSM: Obtain the high-frequency voltage equation of PMSM in the dq coordinate system, neglecting stator resistance and speed-related terms, as follows:

[0047] The expression for the high-frequency response current with respect to the injection voltage is obtained. Inverting the above equation yields:

[0048]

[0049] Integrating the above equation, we obtain the expression for the high-frequency response current:

[0050]

[0051] In the dq coordinate system, neglecting stator resistance and speed-related terms, the PMSM high-frequency voltage equation is:

[0052]

[0053] Finally: Rotational voltage injection is employed, V h ω h Let the amplitude and frequency of the injected voltage be represented as a matrix:

[0054]

[0055] Substituting the above equation into the expression for the high-frequency response current, we obtain the high-frequency response current of the HF-RSVI as follows:

[0056]

[0057] Furthermore, in step S106, iron loss includes hysteresis loss, eddy current loss, and additional loss.

[0058] Furthermore, in step S107, the iron loss generated by the high-frequency injection and the inherent losses of the motor are used to heat the oil pump housing through heat conduction, thereby increasing the oil temperature. The current flowing through the motor windings includes the fundamental wave and high-frequency components, which generate resistance losses, causing the winding temperature to rise. A high-frequency current with a specific amplitude and frequency is injected into the motor stator, generating significant hysteresis losses, eddy current losses, and additional losses in the motor core. These losses are converted into heat energy, causing the core temperature to rise sharply. The motor housing and oil are in close contact, forming a heat conduction path. The oil flowing inside the oil pump housing absorbs heat through convective heat exchange with the heated inner wall of the motor housing, causing the temperature to rise. By adjusting the injection parameters of the high-frequency current, the magnitude of the iron loss can be precisely controlled.

[0059] The principle and effect of this solution are as follows:

[0060] 1. Compared with existing technologies, this method can be widely applied to non-road vehicle electro-hydraulic systems, submersible motor systems, hydraulic servo drive equipment and other fields, providing reliable theoretical and tool support for their startup in low-temperature environments.

[0061] 2. This invention introduces high-frequency injection into the low-temperature starting process of electromechanical fluid. Copper losses are generated by the motor. Due to the significant skin effect of the resistor at high frequencies, the increased resistance promotes winding heating. The high-frequency current increases the iron losses of the motor, causing the motor to heat up and melt solidified grease. The heating power is then dynamically reduced as the temperature rises.

[0062] 3. The voltage amplitude and frequency of high-frequency injection are dynamically changing. When the oil viscosity assessment does not meet the start-up requirements, a high-frequency voltage injection command is generated. By continuously increasing the current frequency until the maximum iron loss operating point is reached, the heating speed is accelerated, and the oil circuit circulation is promoted.

[0063] 4. This invention monitors the oil condition by monitoring the load torque. Therefore, a torque observer is used to observe the load torque. The strong robustness of sliding mode control comes from a large switching gain, but the larger the switching gain, the stronger the chattering in sliding mode control. Using the reaching law method and the quasi-sliding mode method, a novel reaching law and saturation function are designed to effectively suppress chattering while ensuring the robustness of sliding mode control.

[0064] 5. Oil viscosity assessment: Load torque is highly correlated with oil viscosity. When establishing historical data, the actual oil viscosity is measured at each temperature point using a standard viscometer. Simultaneously, current sampling must filter out switching noise and current ripple caused by pulsating torque to improve the accuracy of the torque observer.

[0065] 6. The entire process incorporates multiple protection measures. When the winding overheats, the amplitude automatically decreases. When the load decreases and the oil melts, the injection amplitude is reduced proportionally. After normal operation, the minimum amplitude for maintaining position tracking is used. The system simultaneously monitors parameters such as temperature and response quality to achieve coordinated optimization of injection frequency and amplitude. This control strategy, by dynamically adjusting the high-frequency injection intensity, solves the problem of low-temperature startup while minimizing additional losses during normal operation.

[0066] This method first observes the motor load torque using a sliding film observer algorithm. Then, based on this result, the oil viscosity is analyzed, and the oil viscosity is used to assess the electromechanical-hydraulic system state, determining whether oil pump motor heating control is necessary. If so, high-frequency current is injected to induce copper losses in the motor, while the injection voltage frequency is continuously adjusted to maximize iron losses, thus increasing the motor heat and oil temperature. Once the oil temperature reaches a startable condition, the amplitude and frequency of the injection voltage are gradually reduced, followed by a smooth transition to id=0 control. This method achieves low-temperature cold start of the electromechanical-hydraulic system. Finally, the invention utilizes simulation algorithms to verify the multi-condition adaptability of the torque observation model, accurately identifying the load torque and assessing the oil viscosity. This method not only improves the reliability of low-temperature cold start but also enables coordinated control of the electromechanical-hydraulic system, effectively enhancing the reliability of cold start under low-temperature conditions. Attached Figure Description

[0067] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0068] Figure 1 The diagram illustrates the structure of an electromechanical-hydraulic system for non-road vehicles in the prior art, based on a low-temperature cold start control method using high-frequency injection proposed in this application.

[0069] Figure 2 This paper illustrates a schematic diagram of the structure of a conventional electromechanical-hydraulic system in the prior art of a low-temperature cold start control method based on high-frequency injection proposed in an embodiment of this application.

[0070] Figure 3 The diagram shows a low-temperature cold start control model based on high-frequency injection in a low-temperature cold start control method proposed in an embodiment of this application.

[0071] Figure 4 The flowchart of the low-temperature cold start control method based on high-frequency injection proposed in the embodiments of this application is shown.

[0072] Figure 5 This paper shows a block diagram of the load torque extended sliding mode observer structure in a low-temperature cold start control method based on high-frequency injection proposed in an embodiment of this application.

[0073] Figure 6 The diagram shows the torque curve observed by the extended sliding mode observer for the load torque of high-frequency injection in a low-temperature cold start control method based on high-frequency injection proposed in an embodiment of this application.

[0074] Figure 7 The diagram shows a block diagram of the high-frequency injection control strategy in a low-temperature cold start control method based on high-frequency injection proposed in an embodiment of this application. Detailed Implementation

[0075] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0076] like Figures 1-7 As shown:

[0077] Figure 1 and Figure 2This paper presents the structure of an electromechanical-hydraulic system for off-road vehicles. Such systems typically consist of a motor, gear pump, valve block, and a closed-loop hydraulic medium, exhibiting a compact structure and high power density. However, they face fundamental challenges in low-temperature environments. At temperatures below -20°C, the viscosity of the hydraulic oil increases dramatically, hindering circulation, while the starting load torque of the pump increases by 5 to 10 times. Traditional motor control strategies exhibit serious deficiencies under these conditions. Open-loop V / F control fails to provide sufficient starting torque, leading to motor stall. The lack of auxiliary heating makes it difficult to reduce oil viscosity, significantly reducing system reliability, impacting efficiency, and even causing malfunctions. The system, with its core components—a submersible electric motor (ESP), gear pump, solenoid valve block, and closed-loop hydraulic oil—forms a multi-physical domain coupled energy transfer chain of "electrical energy—mechanical energy—hydraulic energy." As the carrier of system heat dissipation, the thermal properties of the hydraulic oil directly affect the temperature rise and operating efficiency of the entire system. At low temperatures, the hydraulic oil solidifies, exhibiting significant viscous resistance.

[0078] based on Figure 1 and Figure 2 This application proposes a low-temperature cold start control method based on high-frequency injection, such as... Figure 3 , Figure 4 and Figure 5 As shown, it includes the following steps:

[0079] S101: Analyze the motor state and use the sliding diaphragm observer algorithm to observe and feedforward the disturbances caused by the motor state to the load torque and viscosity coefficient;

[0080] Regarding motor status:

[0081] The motor status includes the motor mechanical angle, motor speed, and load torque. Based on the motor status obtained by id=0 control, the given speed is compared with the feedback speed and output to the speed regulator. Given id=0, the current controller controls the d and q axis voltage inputs. The high-frequency pulsating voltage is superimposed on the d axis voltage, causing the d axis current to pulsate. The input d and q axis voltages are transformed by dq\αβ coordinates and then controlled by space vector pulse width modulation to control the three-phase inverter. The DC power supply is output to the motor through the three-phase inverter to control the motor. The position sensor obtains the motor mechanical angle and speed by collecting position signals, the current sensor collects the current and feeds it back to the control loop, and the torque observer receives the current signal and speed signal to observe the load torque.

[0082] In step S101, the oil condition is monitored by monitoring the load torque. Therefore, a torque observer is used to observe the load torque. The strong robustness of sliding mode control comes from the large switching gain. However, the larger the switching gain, the stronger the chattering of sliding mode control. Therefore, a new approaching law and saturation function are designed using the approaching law method and the quasi-sliding mode method. Chattering is effectively suppressed while ensuring the robustness of sliding mode control. However, at the beginning stage of motor operation and when the load torque changes abruptly, the output of the sliding mode controller will have a large instantaneous pulsation. Especially when the load torque changes abruptly, the chattering of the originally stable output will be aggravated again. Therefore, in order to further reduce chattering and improve the robustness of the control system, an observer needs to be designed to observe and feedforward compensate for the disturbances caused by the load torque and viscosity coefficient.

[0083] Analysis of the motor state equations yields:

[0084]

[0085] In the formula For the estimated rotor mechanical angular velocity, Let J be the estimated load torque, J be the moment of inertia, and T be the moment of inertia. e For electromagnetic rotation, B is the momentary viscous friction coefficient, and λ1 is the sliding mode observer gain. For the rotational speed observation error, sat() is a saturation function used to suppress sliding mode chattering, where:

[0086]

[0087] The controller's sampling frequency is much higher than the time it takes for the load torque to change, so the load torque can be considered a constant value within the control cycle, where T... L This represents the load torque.

[0088]

[0089] S102: Based on the motor state, obtain the state equation, construct the extended sliding diaphragm observer equation, derive the observer error equation, and observe the load torque;

[0090] Specifically:

[0091] Ignoring the rotor damping winding, and combining the motion equations in the mathematical model of a permanent magnet synchronous motor, with the motor's mechanical angular velocity and load torque as state variables, electromagnetic torque as input, and the output also being the mechanical angular velocity, the motor's mechanical motion equations are as follows:

[0092]

[0093] Derive the motor state equation from the motor's mechanical motion equation:

[0094]

[0095] In the formula ω m T is the mechanical angular velocity. e T is the electromagnetic torque. L Load torque; J is the moment of inertia.

[0096] Thus, we have the extended sliding mode observer equation:

[0097]

[0098] In the formula This is an estimate of the mechanical angular velocity; Load torque estimate; l is the feedback gain; U is the sliding mode control law.

[0099] Subtracting the two equations above, we have the observer error equation:

[0100]

[0101] In the formula This represents the speed estimation error; This represents the load torque observation error. Let the sliding mode switching surface s = e1, then U = -λsgn(s) - γs, which is derived from the sliding mode attainability condition. We can obtain:

[0102]

[0103] In the formula, γ≥0, therefore we have Pick β≥1.

[0104] When the system trajectory reaches the sliding surface and moves along the sliding surface, That is to say Combined with the observer error equation We can obtain:

[0105]

[0106] After solving, we have:

[0107]

[0108] Where c e Since it is a constant, only when l < 0 can e2 approach zero, thus obtaining the load torque;

[0109] Specifically: The block diagram of the load torque extended sliding mode observer is as follows Figure 5 As shown.

[0110] like Figure 6As shown, to verify the accuracy of torque observation using the proposed method, two sets of experiments were set up. The first set used random load torque to monitor the torque observer's ability to follow the load torque. The second set simulated the actual load torque decreasing as temperature increases, and tested the accuracy of the torque observer in observing the load torque. The simulation results are shown in the figure below, demonstrating that the proposed method can observe load torque well.

[0111] S103: Analyze the obtained load torque. The motor load torque reflects the oil viscosity.

[0112] Specifically:

[0113] The load torque measured in steady state includes the viscous friction torque generated by fluid shear, the mechanical friction torque, and the torque of the pump overcoming pressure to deliver the liquid. The formula is as follows:

[0114] T e =T vis +T m +T l =c*μ(T)*ω+T m +T l

[0115]

[0116] In the formula T e For the load torque, T vis The viscous frictional torque is generated by fluid shearing, c is a constant coefficient, and μ(T) is the dynamic viscosity of the oil at temperature T. m T is the mechanical friction torque. l The torque required by the pump to overcome pressure and deliver liquid can be calculated using a formula. Given a specific temperature, speed, and measured torque value under those operating conditions, the corresponding viscosity estimate can be calculated. Real-time torque signals at a stable speed are simultaneously acquired, and after necessary processing, the actual measured torque value at that speed is obtained. Based on the current speed and load torque, the estimated oil viscosity at that temperature and speed is calculated.

[0117] S104: Based on the oil viscosity results, high-frequency current injection is performed at low temperature. The injection frequency is controlled to enhance the skin effect and the current amplitude is controlled to utilize copper loss and high-frequency magnetic field eddy current effect to assist in heat generation.

[0118] Specifically:

[0119] For the high-frequency voltage injection method (HFVI), since the injected signal is a high-frequency signal, the high-frequency inductive reactance is much greater than the stator resistance. If only iron losses are analyzed, the voltage drop across the stator resistance can be ignored. Furthermore, since the back EMF exists only in the fundamental frequency component of the voltage, the high-frequency voltage equation for the PMSM can disregard the back EMF. The high-frequency voltage equation for the PMSM can be expressed as:

[0120]

[0121] In the formula, L0=(L d +L q ) / 2、L1=(L d -L q ) / 2, L d L q These are the d-axis and q-axis stator inductances, respectively.

[0122] To obtain the expression for the high-frequency response current with respect to the injection voltage, the inverse of equation (1) above can be obtained as follows:

[0123]

[0124] Integrating equation (2) above, we obtain the expression for the high-frequency response current as follows:

[0125]

[0126] In the dq coordinate system, neglecting stator resistance and speed-related terms, the PMSM high-frequency voltage equation is:

[0127]

[0128] Using rotating voltage injection (4), V h ω h Let the amplitude and frequency of the injected voltage be represented as a matrix:

[0129]

[0130] Substituting the above equation into the expression for the high-frequency response current (3), we obtain the high-frequency response current of the HF-RSVI as follows:

[0131]

[0132] S105: After high-frequency current injection, the motor generates copper losses. At low temperatures, the winding resistance decreases. High-frequency current causes the skin effect, which increases resistance and copper losses.

[0133] Specifically:

[0134] As shown in the copper loss calculation formula, increasing the current can increase the copper loss of the motor. Due to the skin effect, the high-frequency current increases the effective resistance when using the high-frequency voltage injection method (HFVI). The copper loss calculation formula is as follows:

[0135]

[0136] In the formula I d I is the d-axis current. qR is the q-axis current, m is the number of motor phases, and R is the q-axis current. s Let I be the resistance of the motor at temperature θ. dh For the d-axis current injected at high frequency, I qh For the q-axis current injected at high frequency, R a For the motor at temperature θ a The resistance at time θ is the current temperature.

[0137] S106: After high-frequency current injection, the motor generates iron loss. The maximum iron loss point is found by continuously increasing the frequency of the high-frequency current.

[0138] Specifically:

[0139] When a high-frequency current is injected into the stator winding of a motor, a rapidly changing magnetic field is generated in the motor core. The peak point can be found by continuously increasing the injection frequency f while maintaining a fixed high-frequency current amplitude, and measuring or calculating the iron loss at each frequency point. A frequency converter or power amplifier capable of outputting a precise and controllable high-frequency sinusoidal current should be used. The platform must have the capability of continuously adjustable frequency and precise current amplitude control. For rotating motors, the rotor should be reliably stalled to prevent rotation. This is to eliminate the influence of mechanical losses and output power, focusing on iron loss measurement. A fixed high-frequency current amplitude should be selected. This amplitude should be large enough to produce measurable iron loss, but not so large as to cause excessive heating or saturation. The amplitude should remain constant throughout the frequency sweep. At each frequency point, wait for the system to reach an electrically stable state, and measure or estimate the AC resistance and iron loss of the stator winding at that frequency. Iron loss is positively correlated with the frequency of the high-frequency injected current; as the frequency increases, the iron loss also increases. The maximum iron loss point is found when the maximum value is reached, as shown in the following formula:

[0140]

[0141] In the formula P eddy For iron loss, f h Let B be the current frequency and B be the magnetic flux density.

[0142] S107: Motor copper loss causes winding temperature to rise. After high-frequency current injection, the motor generates iron loss, which causes the iron core to heat up. The heat is conducted to the casing and the oil temperature is increased by heating the oil pump casing.

[0143] Specifically:

[0144] Utilizing the inherent losses of the motor itself and the iron losses generated by high-frequency injection, the oil pump housing is heated through heat conduction, thereby increasing the oil temperature. The current flowing through the motor windings, including the fundamental wave and high-frequency components, generates resistive losses, leading to an increase in winding temperature. Injecting a high-frequency current of specific amplitude and frequency into the motor stator generates significant hysteresis losses, eddy current losses, and additional losses in the motor core. Almost all of these losses are converted into heat energy, causing a sharp rise in core temperature. The motor housing and oil are in close contact, forming a heat conduction path. The oil flowing inside the oil pump housing absorbs heat through convective heat exchange with the heated inner wall of the motor housing, increasing its temperature. By adjusting the injection parameters of the high-frequency current, the magnitude of iron losses can be precisely controlled, thereby controlling the heat conducted to the oil pump housing, ultimately achieving an active and controllable increase in oil temperature.

[0145] S108: Based on the final oil viscosity result, the electromagnetic heating viscosity decreases and the load is reduced, so the injection voltage amplitude is gradually reduced. After the normal start-up conditions are met, the control switches to normal id=0.

[0146] Heat is conducted to the oil pump housing, heating the circulating oil. The increased oil temperature leads to a significant decrease in oil viscosity, reducing load torque and power demand. This reduces the electromagnetic torque required for the motor to maintain the same speed, consequently lowering the required stator current amplitude. As viscosity decreases and the load lightens, the amplitude of the high-frequency injection voltage used for heating is gradually reduced to decrease heat generation, prevent oil overheating, and achieve energy savings. When the oil viscosity drops to a level permissible for normal startup, the normal startup conditions are deemed met. High-frequency injection heating is stopped, and the system smoothly transitions to normal motor operation in vector control mode (id=0).

[0147] High-frequency injection control strategies such as Figure 7 As shown, the system first observes the load torque and then evaluates the viscosity using this load torque. A sliding mode observer is constructed based on the motor's motion equations, and the load torque is estimated in real time using current and speed information. The observer accurately captures load changes through a specially designed speed error processing mechanism. In oil pump applications, this observed value directly reflects the oil's viscosity state and is a key criterion for switching control strategies.

[0148] If the starting requirements are met, the oil pump motor starts. If not, the controller injects a specific high-frequency voltage signal into the motor stator windings. This signal is in the form of a rotating voltage wave, with a frequency set in the range of 500-1000Hz, much higher than the motor's fundamental frequency. The amplitude of the injected signal is dynamically adjusted according to the system status: the maximum amplitude is used during low-temperature startup, and reduced to the minimum necessary value during normal operation. This high-frequency signal is superimposed on the fundamental control voltage and applied to the motor windings. After a period of injection, copper and iron losses are calculated. If the maximum iron loss operating point is reached, a maximum iron loss high-frequency current command is generated; otherwise, the current frequency is increased to reach the maximum iron loss operating point. As the temperature continues to rise, the system continues to monitor torque and viscosity until the requirements for starting the oil pump motor are met.

[0149] This invention discloses a low-temperature cold start control method based on high-frequency injection. The method first observes the motor load torque using a sliding film observer algorithm. Then, based on this result, the oil viscosity is analyzed, and the oil viscosity is used to assess the state of the electromechanical-hydraulic system to determine whether oil pump motor heating control is necessary. If so, high-frequency current is injected to induce copper losses in the motor, while the injection voltage frequency is continuously adjusted to maximize iron losses, thereby increasing the motor heat and raising the oil temperature. Once the oil temperature reaches a startable condition, the amplitude and frequency of the injection voltage are gradually reduced, followed by a smooth transition to id=0 control. This method achieves low-temperature cold start of the electromechanical-hydraulic system. Finally, the invention utilizes simulation algorithms to verify the multi-condition adaptability of the torque observation model, accurately identifying the load torque and assessing the oil viscosity. This method not only improves the reliability of low-temperature cold start but also enables coordinated control of the electromechanical-hydraulic system, effectively enhancing the reliability of cold start under low-temperature conditions.

[0150] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any indirect modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A low-temperature cold start control method based on high-frequency injection, characterized in that, Includes the following steps: S101: Analyze the motor state and use the sliding mode observer algorithm to observe and feedforward the disturbances caused by the motor state to the load torque and viscosity coefficient; S102: Based on the motor state, obtain the state equation, construct the extended sliding mode observer equation, derive the observer error equation, and observe the load torque; The state equation is as follows: Motor state equations: ; In the formula For the estimated rotor mechanical angular velocity, For the estimated load torque, For rotational inertia, For electromagnetic rotation, B is the coefficient of viscous friction. Sliding mode observer gain, For speed observation error, ( ) is a saturation function used to suppress sliding mode chattering, where: ； The motor controller's sampling frequency is higher than the load torque change time relative to the load torque change time, assuming the load torque is a constant value within the control cycle, i.e.: ; Combining the motion equations in the mathematical model of the permanent magnet synchronous motor, with the motor's mechanical angular velocity... and load torque Electromagnetic torque is a state variable. The input is mechanical angular velocity, and the output is also mechanical angular velocity. The state equations are as follows: ； Based on the state equations, the extended sliding mode observer equations are constructed as follows: ； In the formula This is an estimate of the mechanical angular velocity; Estimated load torque; Here, U is the feedback gain; U is the sliding mode control law. Subtracting the two equations from the above gives the observer error equation: The observer error equation is as follows: ; In the formula This represents the speed estimation error; To account for the load torque observation error, let the sliding mode switching surface... ,have The conditions that can be met by sliding mode We can obtain: ； In the formula Therefore, ,Pick , ; Combining the observer error equation, we obtain ; in For a constant, only if it satisfies Only then can we guarantee Approaching zero, thus yielding the load torque; S103: Analyze the obtained load torque. The motor load torque reflects the oil viscosity. A historical benchmark database is established, which contains known viscosity samples. Given a specific temperature, rotational speed, and torque value measured under that operating condition, the corresponding viscosity estimate can be queried or calculated. Real-time torque signals at stable rotational speeds are collected synchronously, and the collected real-time torque signals are processed to obtain the actual measured torque value at that rotational speed. In the established historical database, based on the current rotational speed and load torque, the expected benchmark torque value at that temperature and rotational speed is found or interpolated and calculated. The current oil viscosity estimate is then calculated and output using the benchmark torque value. S104: Based on the oil viscosity results, high-frequency current injection is performed at low temperature. The injection frequency is controlled to enhance the skin effect and the current amplitude is controlled to utilize copper loss and high-frequency magnetic field eddy current effect to assist in heat generation. S105: After high-frequency current injection, the motor generates copper losses. At low temperatures, the winding resistance decreases. High-frequency current causes the skin effect, which increases resistance and copper losses. S106: After high-frequency current injection, the motor generates iron loss. The maximum iron loss point is found by continuously increasing the frequency of the high-frequency current. S107: Motor copper loss causes winding temperature to rise. After high-frequency current injection, the motor generates iron loss, which causes the iron core to heat up. The heat is conducted to the casing and the oil temperature is increased by heating the oil pump casing. S108: Based on the final oil viscosity result, the electromagnetic heating viscosity decreases and the load is reduced, so the injection voltage amplitude is gradually reduced. After the normal start-up conditions are met, the control switches to normal id=0.

2. The low-temperature cold start control method based on high-frequency injection according to claim 1, characterized in that, Motor status includes motor mechanical angle, speed, and load torque. Based on the motor status obtained from id=0 control, the given speed is compared with the feedback speed and output to the speed regulator. Given id=0, the current controller controls the d-axis and q-axis voltage inputs. A high-frequency pulsating voltage is superimposed on the d-axis voltage, causing d-axis current pulsation. The input d-axis and q-axis voltages are then processed... Coordinate transformation is performed, and then the three-phase inverter is controlled by space vector pulse width modulation. The DC power supply is output to the motor through the three-phase inverter to control the motor. The position sensor obtains the mechanical angle and speed of the motor by collecting position signals, the current sensor collects the current and feeds it back to the control loop, and the torque observer receives the current signal and speed signal to observe the load torque.

3. The low-temperature cold start control method based on high-frequency injection according to claim 2, characterized in that, In step S014, the high-frequency voltage equation of PMSM in the dq coordinate system is first required; secondly, the expression of the high-frequency response current is obtained based on the high-frequency voltage equation of PMSM; then, rotating voltage injection is used to represent the injection voltage in matrix form and combine it with the expression of the high-frequency response current to obtain the high-frequency response current of HF-RSVI and the current response demodulation.

4. The low-temperature cold start control method based on high-frequency injection according to claim 3, characterized in that, The high-frequency response current of HF-RSVI is obtained as follows: First: The high-frequency voltage equation of PMSM can be expressed as: ； In the formula, , , These are the d-axis and q-axis stator inductances, respectively. Then: Inverse and integral the high-frequency voltage equation of PMSM: Obtain the high-frequency voltage equation of PMSM in the dq coordinate system, neglecting stator resistance and speed-related terms, as follows: The expression for the high-frequency response current with respect to the injection voltage is obtained. Inverting the above equation yields: ； Integrating the above equation, we obtain the expression for the high-frequency response current: ； In the dq coordinate system, neglecting stator resistance and speed-related terms, the PMSM high-frequency voltage equation is: ； Finally: Rotational voltage injection is employed. , Let the amplitude and frequency of the injected voltage be represented as a matrix: ； Substituting the above equation into the expression for the high-frequency response current, we obtain the high-frequency response current of the HF-RSVI as follows: 。 5. The low-temperature cold start control method based on high-frequency injection according to claim 4, characterized in that, In step S106, iron loss includes hysteresis loss, eddy current loss, and additional loss.

6. The low-temperature cold start control method based on high-frequency injection according to claim 5, characterized in that, In step S107, the oil temperature is increased by heating the oil pump housing through heat conduction, utilizing the motor's own losses and the iron losses generated by high-frequency injection. The current flowing through the motor windings, including the fundamental wave and high-frequency components, generates resistance losses, causing the winding temperature to rise. High-frequency current with a specific amplitude and frequency is injected into the motor stator, generating significant hysteresis losses, eddy current losses, and additional losses in the motor core. These losses are converted into heat energy, causing the core temperature to rise sharply. The motor housing and oil are in close contact, forming a heat conduction path. The oil flowing inside the oil pump housing absorbs heat through convective heat exchange with the heated inner wall of the motor housing, causing the temperature to rise. By adjusting the injection parameters of the high-frequency current, the magnitude of iron loss can be precisely controlled.

Images