Method for estimating the insulation resistance of a high-voltage circuit in an electric or hybrid motor vehicle

Abstract

A method for determining the insulation resistance of a high-voltage network in an electric or hybrid vehicle, comprising: providing a controllable continuous voltage source connected to a single first terminal of a vehicle body and a high-voltage battery of the vehicle; a first resistor connected in series with the source between the single first terminal and the vehicle body; a second resistor connected in series between the first resistor and the source; applying continuous voltage setpoint values ​​between the vehicle body and the single first terminal; acquiring a measurement signal of the voltage across the second resistor for each setpoint value; performing adaptive filtering on the signal; recursively estimating a vector of filter transfer function coefficients; providing an update to the filter coefficients; and determining the insulation resistance based on the estimate.

Description

[0001] This invention relates to the field of electric or hybrid vehicles. More specifically, it relates to a method for determining the insulation resistance between a point on the high-voltage circuit of an electric or hybrid vehicle, including a high-voltage battery, and the vehicle's ground terminal.

[0002] In electric or hybrid motor vehicles with a traction system equipped with at least one electric motor, the electric motor provides the motor torque required to drive the vehicle. For this purpose, electricity is supplied to the electric motor from a high-voltage battery. The required voltage level reaches several hundred volts, typically around 400 volts. This battery also has a large capacity to increase the vehicle's range in electric mode. Many technical reasons specific to automotive applications necessitate the use of insulation between the vehicle body or mechanical grounding terminal (formed by the vehicle's metal chassis and metal body, and therefore accessible to the user) and the battery's potential. Therefore, any part of the vehicle electrically connected to the battery must be insulated from the grounding terminal. This insulation is achieved by using electrically insulating materials. However, over time, the insulation may deteriorate, thus placing the vehicle's grounding terminal at a potential that poses a danger to vehicle passengers or anyone in contact with the vehicle.

[0003] This is why it is essential to check the insulation resistance between any point on the high-voltage circuit and the vehicle's grounding terminal to prevent potential electric shock to vehicle passengers or anyone in contact with the vehicle. Specifically, this monitoring allows for the correction of a first insulation fault before a second one occurs. In essence, only a double fault will cause a short circuit that could lead to vehicle malfunction.

[0004] The insulation resistance can be measured or estimated using physical methods.

[0005] Measuring insulation resistance physically requires injecting high voltage and high current into the insulation point being measured. However, it is necessary to be able to know the insulation resistance between any point on the vehicle's high-voltage circuit and the vehicle's grounding terminal at any time, while still allowing the system to continue operating, in other words, without causing traction or load interruption.

[0006] Therefore, when inspecting the insulation of the high-voltage network of electric and hybrid vehicles during operation, it is preferable to obtain the insulation resistance by estimation.

[0007] The estimate must be accurate enough, but it must not be overestimated in order to allow for reliable detection of any insulation faults; nor should it be underestimated in order to avoid any false detections that could lead to vehicle malfunctions.

[0008] Devices for detecting insulation faults in electric or hybrid vehicles are known in the prior art. These devices are based on resistance measurement circuits, in which leakage current is measured using a voltage divider bridge formed by multiple resistors connected between the battery terminals. A disadvantage of this circuit is that it requires connection to both terminals of the battery, which complicates its integration within the vehicle.

[0009] Document FR 3037406 discloses a circuit for detecting electrical insulation faults between a high-voltage battery of an electric vehicle and the vehicle body (which forms the vehicle's electrical ground). The detection circuit is electrically connected to a single first terminal of the battery, such as the negative terminal of the battery. An insulation fault typically manifests as a point on the battery (e.g., a point between two adjacent battery cells) having an insulation resistance below a safe threshold with respect to the battery body. Therefore, a potentially dangerous leakage current flows from this point on the battery through this insulation resistance to the battery body. The detection circuit according to the aforementioned document includes a controllable DC voltage source connected to both the battery body and the single first terminal of the battery, allowing various voltage values ​​to be applied between the battery body and the single first terminal. The detection circuit also includes a device for measuring a current that flows through the single first terminal of the battery and out through a point on the battery, representing an insulation fault, which then flows through the insulation resistance to the battery body. Thus, the voltage applied by the controllable voltage source generates a current flowing through the measuring device and the insulation resistance. In this way, for each voltage value applied by a controllable voltage source, a measured value of the current is obtained, thereby allowing the calculation of the insulation resistance value between the point in question on the battery and the vehicle body.

[0010] However, calculating insulation resistance values ​​using this method is relatively time-consuming. Specifically, it requires biasing the individual first terminals of the vehicle body and battery at different potentials for a certain period to reach a steady state where leakage current measurements can be performed. The time required to achieve this steady state may hinder the implementation of reliable strategies for inspecting the insulation of the vehicle's high-voltage network.

[0011] Therefore, the object of the present invention is to overcome this limitation, at least in part.

[0012] Therefore, the present invention relates to a method for determining the insulation resistance between a point on a high-voltage battery of an electric or hybrid vehicle and the vehicle body, the method comprising the following steps:

[0013] Provides a controllable DC voltage source connected to a single first terminal of the vehicle body and the battery.

[0014] A first resistor, connected in series with the voltage source, is provided between the single first terminal and the vehicle body. This first resistor is capable of limiting the current flowing into the single first terminal of the battery.

[0015] A second resistor is provided to be connected in series between the first resistor and the voltage source.

[0016] Provide a measuring device capable of measuring the voltage across the second resistor.

[0017] Different continuous voltage setpoint values ​​are applied between the vehicle body and a single first terminal of the battery by the voltage source.

[0018] The measuring device acquires a voltage measurement signal, which represents the measured voltage across the second resistor for each continuously applied voltage setpoint value.

[0019] The value of the insulation resistance is calculated based on the voltage measurement signal.

[0020] The method is characterized by implementing an adaptive filtering step for the voltage measurement signal, and includes the following steps: recursively estimating a vector of transfer function coefficients of the adaptive filter, providing an update of the filter coefficients, and calculating the insulation resistance value based on the estimated value.

[0021] The method of this invention enables optimization of the convergence speed of the method for determining insulation resistance without increasing computational complexity.

[0022] Advantageously, the value of this insulation resistance is calculated using the following formula:

[0023]

[0024] Where Ri is the value of the insulation resistance, Rd is the value of the first resistor, and Rm is the value of the second resistor, and It is an estimate of the vector of transfer function coefficients of the adaptive filter.

[0025] Advantageously, the vector of transfer function coefficients of the adaptive filter is estimated by considering the transfer functions of the input signal u(k) and the output signal y(k), which are defined as follows:

[0026] u(k)=U d2 (k)-U d1 (k)=ΔU d (k)

[0027] y(k)=U m2 (k)-U m1 (k)=ΔU m (k)

[0028] Among them, U m1 (k) and U m2(k) respectively correspond to these voltage setpoint values ​​U applied continuously by the voltage source. d1 (k) and U d2 (k) is the voltage value measured across the two ends of the second resistor.

[0029] Advantageously, the input signal to the filter transfer function is replaced with an input signal defined as follows:

[0030]

[0031] Among them, U bat1 and U bat2 These correspond to the total battery voltage values ​​considered for each voltage setpoint value continuously applied by the voltage source, in order to account for the change in total battery voltage during each iteration of the applied voltage setpoint value.

[0032] Preferably, the method includes the step of calculating the position of the insulation resistance relative to a single first terminal of the battery based on the estimated value.

[0033] Advantageously, the vector of transfer function coefficients of the adaptive filter is estimated by considering the transfer functions of the input signal u(k) and the output signal y(k), which are defined as follows:

[0034] u(k)=U bat ·(U m2 -U m1 )

[0035] y(k)=U m1 U d2 -U d1 U m2

[0036] Among them, U m1 and U m2 These correspond to the voltage setpoint values ​​U that are continuously applied by the voltage source. d1 and U d2 The voltage value measured across the second resistor, and U bat This is the total battery voltage.

[0037] Advantageously, the input signal to the filter transfer function is replaced with an input signal defined as follows:

[0038] u(k)=U m2 ·U bat1 -U m1 ·U bat2

[0039] Among them, U bat1 and U bat2These correspond to the total battery voltage values ​​considered for each voltage setpoint value continuously applied by the voltage source, in order to account for the change in total battery voltage during each iteration of the applied voltage setpoint value.

[0040] Advantageously, this filtering step is performed using a recursive least squares algorithm.

[0041] Advantageously, the method includes using a single filter setting parameter, which represents the forgetting factor of the filter and consists of real coefficients with values ​​between 0 and 1.

[0042] The present invention also relates to an apparatus for determining the insulation resistance between a point on a high-voltage battery (1) of an electric or hybrid vehicle and the vehicle body, the apparatus comprising a detection circuit for detecting an insulation fault between the battery and the vehicle body, the detection circuit comprising: a controllable DC voltage source capable of applying a voltage setpoint between a single first terminal of the battery and the vehicle body; a first resistor connected in series with the voltage source between the single first terminal and the vehicle body, the first resistor being capable of limiting the current flowing through the single first terminal of the battery; and a second resistor connected in series between the first resistor and the voltage source; and so on. The device includes a measuring apparatus capable of measuring the voltage across the second resistor; the apparatus includes a control unit capable of applying different continuous voltage setpoint values ​​between a vehicle body and a single first terminal of the battery via the voltage source; capable of acquiring a voltage measurement signal via the measuring apparatus, the voltage measurement signal representing the measured voltage across the second resistor for each continuously applied voltage setpoint value; and capable of calculating the value of the insulation resistance based on the voltage measurement signal. The control unit is characterized by including an adaptive filtering module for implementing adaptive filtering processing according to the above method.

[0043] Other features and advantages of the invention will become clearer from the following description, which is given as an illustrative rather than limiting example and with reference to the single accompanying drawing:

[0044] [ Figure 1 This schematically illustrates a circuit used to detect insulation faults in the high-voltage battery of an electric or hybrid motor vehicle.

[0045] [ Figure 2 ]Schematic illustration Figure 1 Insulation faults in high-voltage batteries.

[0046] The following description relates to an exemplary embodiment of the invention applied to high-voltage power grids in electric or hybrid motor vehicles. References Figure 1Electric or hybrid motor vehicles include a rechargeable high-voltage battery 1 as an onboard high-voltage power source. Battery 1 includes two terminals, a positive terminal HV+ and a negative terminal HV-. Battery 1 consists of a group of battery cells connected in series between the two terminals HV- and HV+. Battery 1 is designed to deliver a DC voltage U. bat The value of this DC voltage remains constant over time. In this example, the battery voltage U bat It equals 400V.

[0047] Connected to the battery terminals HV+ and HV- are electrical loads (not shown), including the inverter and the electric motor used to propel the vehicle. The inverter converts the battery voltage U... bat The voltage is converted to AC power supply voltage for electric motors. The vehicle also includes a body 2 formed by the vehicle chassis and body, which are typically made of metal. This body constitutes the vehicle's electrical grounding terminal, to which charge can flow in the event of an insulation failure at a point on the vehicle's high-voltage battery.

[0048] The insulation fault here means that there is an abnormally low-resistance electrical contact between the vehicle body 2 and a potential point on the battery (for example, one of the battery terminals HV+ and HV-). Figure 1 Resistances 31 and 32 are shown between the vehicle body 2 and each of the various terminals HV+ and HV- of the battery, with values ​​denoted by Ri+ and Ri-, respectively. A resistance value is considered low if it is below or equal to a predefined safety threshold (e.g., 100 kΩ). Typically, in the absence of insulation failure, the resistance between one side of the vehicle body 2 and the other side's terminals HV+ and HV-, and more generally, the resistance between the vehicle body and any potential point on the battery 1, is greater than 100 kΩ or 1 MΩ. Alternatively, this resistor can be modeled as an infinite value. Due to this high resistance value, there is no potentially dangerous leakage current flow between the battery 1 and the vehicle body 2.

[0049] Figure 2 This illustrates a single insulation fault between point 3 on battery 1 and vehicle body 2. This insulation fault manifests here as insulation resistance 30, which connects point 3 and vehicle body 2 and has a connection to R. i The value represents a value below the safety threshold. A potentially dangerous leakage current then flows from battery 1 to vehicle body 2 through resistor 30. This leakage current is undesirable and could endanger vehicle occupants who may have direct contact with vehicle body 2. For example, point 3 is located between two adjacent battery cells of battery 1. Battery 1 can then be likened to two DC voltage sources 11 and 12 on either side of point 3, connected in series between terminals HV+ and HV-. Sources 11 and 12 deliver a voltage (1-α)*U between their terminals respectively. bat and α*Ubat Wherein, the coefficient α is a real number in the range [0; 1]. Knowing the coefficient α, the location of the insulation fault in battery 1 can be determined. Therefore, the insulation resistance indicating the insulation state of the battery is located at position α relative to the battery terminal HV- and position (1-α) relative to the battery terminal HV+.

[0050] like Figure 1 and Figure 2 As shown, the vehicle further includes a circuit 4 for detecting such an insulation fault between the vehicle's high-voltage battery 1 and the vehicle's body 2. The detection circuit 4 is electrically connected between the body 2 and a single terminal (here, terminal HV-) of the battery 1. In other words, the terminal of the detection circuit 4 is connected on one side to the body 2 (which constitutes the vehicle's ground terminal) and on the other side to terminal HV- of the vehicle's high-voltage battery 1.

[0051] The detection circuit 4 includes a controllable DC voltage source 41, which is capable of adjusting the non-zero voltage setpoint U according to the control signal received from the control unit 46. d The circuit applied between terminal HV- and the vehicle body is used to generate a potential difference across the insulation resistance 30 if necessary, and thus to allow current i to flow. d The current flows through the insulation resistor. Voltage setpoint value U d Preferably, the voltage is below or equal to 60V, for example, between 0V and 24V.

[0052] The detection circuit 4 also includes a first resistor 42 (referred to as the current-limiting resistor 42), which is connected in series with the DC voltage source 41 between the battery terminal HV- and the vehicle body. More precisely, the current-limiting resistor 42 is connected between the terminal HV- and the DC voltage source 41. This current-limiting resistor 42 ensures better insulation between the battery 1 and the rest of the detection circuit 4, thereby preventing current i d An excessively high value poses a danger to the user. (By R) d The value of the current-limiting resistor 42 is, for example, chosen to be as low as possible to facilitate the flow of current i. d The measurement was selected to be high enough not to damage the electrical insulation of the detection circuit 4. d The value will preferably be selected as greater than the value of the abnormal insulation resistance R. i High (e.g., five times higher, or even ten times higher), for example, for a voltage of 400V, the value Ri of this abnormal insulation resistance is on the order of 100kΩ, resulting in a maximum acceptable current of 4mA (the highest acceptable current generally accepted for personal safety is on the order of 10mA). For example, the value R of the current-limiting resistor... d Therefore, it equals 500kΩ.

[0053] The detection circuit 4 also includes a device 43 for measuring the current flowing through the battery terminal HV- and the insulation resistance 30. The measuring device 43 includes a second resistor 44 (referred to as the measuring resistor 44), the value of which is determined by R. m This indicates that the measuring resistor and the capacitor 45 with capacitance Cm are connected in parallel between the DC voltage source 41 and the current-limiting resistor 42. In other words, the DC voltage source 41, the measuring resistor 44, and the current-limiting resistor 42 are connected in series between the vehicle body 2 and the HV- terminal of the battery. When the DC voltage source 41 sets the voltage to the set point U... d When applied to the detection circuit 4, the measuring device 44 is able to measure the voltage U across the measuring resistor 44. m Because the value R of resistor 44 is being measured. m Since it is known, the voltage U across the measuring resistor 44 is measured. m This allows the current i to be automatically derived from it. d The current flows through the current-limiting resistor 42 to enter through the battery terminal HV-, and then through the insulation resistor 30 to the vehicle body 2.

[0054] Therefore, the voltage setpoint U d The application generates a current i through the measuring device 43 and the current-limiting resistor 42. d .

[0055] Therefore, in order to estimate the value R of the insulation resistance 30... i The control unit 46 is capable of controlling the continuous application of multiple different voltage setpoints U by the DC voltage source 41. d The value can then be used to obtain the corresponding current i measured by the measuring device 43 for each voltage setpoint value applied by the DC voltage source 41. d The value of i. Then, the control unit can, based on the acquired current i d Value and applied voltage setpoint U d The value of R is automatically calculated to be the insulation resistance 30 associated with the insulation fault located at point 3 on the battery. i .

[0056] For example, DC voltage source 41 is controlled to continuously apply two different voltage setpoint values ​​(by U). d1 and U d2 (Represented). These voltage setpoint values ​​U d1 and U d2 Each of them corresponds to the measuring resistor R. m The measured values ​​of the voltage across the two ends (respectively by U) m1 and U mm2 express).

[0057] Therefore, the value of insulation resistance 30 is R iThe position α of the insulation resistance relative to the battery terminal HV- is calculated according to the following equation, where the total battery voltage U is assumed. bat The applied voltage setpoint value is constant in each iteration:

[0058]

[0059]

[0060] If voltage U needs to be considered bat With the change, these equations become:

[0061]

[0062]

[0063] U batl and U bat2 These are two voltage setpoints U applied consecutively. d1 and U d2 The battery voltage should be considered at that time.

[0064] Due to the measurement resistor R m Voltage U across the terminals m Measurements are typically noisy, therefore equations (1) and (2) or (3) and (4) are not used directly. According to the invention, these equations are processed by two corresponding adaptive algorithms operating on a known recursive least squares (RLS) model, which are capable of providing two quantities R... i And a stable estimate of each of α. Therefore, the control unit of the detection circuit is able to periodically provide a reliable indication of the electrical insulation status of the traction system to the battery management system (BMS), thereby allowing necessary measures to be taken when an insulation fault is observed.

[0065] An exemplary embodiment of such adaptive filtering, implemented by recursive least squares (RLS), will now be described in more detail. This adaptive filtering is used to estimate insulation resistance on one hand and to estimate the location of insulation faults in the battery on the other.

[0066] First, recall the equations for implementing the RLS algorithm. We will only consider the discrete case where the signal is sampled, where index k represents the index of the current value of the quantity.

[0067] Typically, this algorithm can be used to estimate the coefficients a of the discrete transfer function H(z) of the input signal u(k) and the output signal y(k). i and b i :

[0068]

[0069] Therefore, this is a problem of finding the "optimal" filter, that is, when the input is a given sequence u(k), the filter can obtain a response y(k) at the output that is "closest" to the desired response. This adaptive filtering involves recursive updates of the filter coefficients. Thus, the algorithm starts with predetermined initial conditions and recursively modifies the filter coefficients to adapt to the process.

[0070] The main steps of each RLS algorithm are:

[0071] 1. Initialization steps: 0 < λ ≤ 1; C(O) = C0; θ(0) = 0;

[0072] C(k) is the covariance matrix of the input variables of the algorithm, which is equal to C0 at the initial time k = 0. λ is the forgetting factor parameter. θ(k) is the vector of filter coefficients to be estimated, which is set to 0 at the initial time k = 0.

[0073] 2. Prediction steps:

[0074] This corresponds to the prediction error of the standard to be minimized. In this case, the goal is to minimize the difference between the current value y(k) of the signal generated at the algorithm's output and the result of filtering the past values ​​of the signal using a transfer function with estimated coefficients θ(k-1). T (k) denotes the transpose of the vector X(k) as defined below:

[0075] X(k)=[u(k)u(k-1)...u(k-nb)-y(k-1)...-y(k-na)] T

[0076] and

[0077] θ(k)=[b0 b1 ... b nb a1 ... a na ] T

[0078] 3. Steps for calculating the optimal gain:

[0079] L(k)=C(k-1)·X(k)·[λ+X T (k)·C(k-1)·X(k)] -1

[0080] 4. Steps to update the estimated value of the filter coefficient vector θ(k):

[0081]

[0082] This adaptation of the filter coefficients is performed based on the prediction error, which makes it possible to compute updates to the filter coefficients.

[0083] 5. Steps to update the covariance matrix C(k):

[0084]

[0085] In a steady state where the target impedance is purely resistive, the equation (5) defined above becomes:

[0086] H(z) = b0

[0087] In other words, the goal of the adaptive algorithm is to provide estimates of the vector of filter coefficients in two different implementations of the adaptive algorithm, from which the value R can be derived. i And α, the estimate here is determined by a single parameter. The parameter is composed of the static gain b0 of the transfer function H(z) in equation (5).

[0088] The first implementation method is used to estimate the value R of the insulation resistance. i As mentioned above, depending on whether the change in battery voltage Ubat is considered, two cases may occur. Therefore, in the first case where the change in battery voltage Ubat is ignored, the static gain b0 of the transfer function H(z) of equation (5) is evaluated by considering the transfer functions H(z) of the input signal u(k) and the output signal y(k), which are defined as follows:

[0089] u(k)=U d2 (k)-U d1 (k)=ΔU d (k)

[0090] y(k)=U m2 (k)-U m1 (k)=ΔU m (k)

[0091] Therefore, it is used to update the estimate. The equation is:

[0092]

[0093] in:

[0094]

[0095]

[0096] After the algorithm converges, we get:

[0097]

[0098] Based on equation (1), the estimated insulation resistance R i The calculation is as follows (7):

[0099]

[0100] Considering battery voltage U bat In the second case of the change, the static gain b0 of the transfer function H(z) in equation (5) is evaluated in the same way as in the first case, but the transfer function H(z) takes into account the following input signal:

[0101]

[0102] A second implementation of the adaptive algorithm is used to estimate a value α, which corresponds to the location of the insulation fault relative to the battery terminal HV-. Similar to estimating the insulation resistance, this depends on whether the battery voltage U is considered. bat The change in voltage Ubat can lead to two scenarios. Therefore, in the first case where the change in battery voltage Ubat is ignored, steps 1 to 5 of the adaptive algorithm are implemented using the input signal u(k) and output signal y(k) of the transfer function H(z). These input and output signals are defined as follows:

[0103] u(k)=U bat ·(U m2 -U m1 )and

[0104] y(k)=U m1 U d2 -U d1 U m2

[0105] After the algorithm converges, the vector of filter coefficients The estimate provides an estimate of the value α.

[0106] Considering battery voltage U bat In this case of variation, the input signal u(k) and output signal y(k) of the transfer function H(z) are used to perform estimation. In steps 1 to 5 of the adaptive algorithm, the input signal and the output signal are defined as follows:

[0107] u(k)=U m2 ·U bat1 -U m1 ·U bat2 and

[0108] y(k)=U m1 U d2 -Ud1 U m2

[0109] Whether estimating insulation resistance or insulation fault location, implementing the RLS-type adaptive algorithm has the advantage of having only one setting parameter. This setting parameter, representing the forgetting factor λ, is defined in the equation above. The value of this parameter is between 0 and 1, allowing for the assignment of greater or less importance to earlier samples in the covariance matrix. Therefore, parameterization of this parameter is essential to optimize the algorithm's convergence speed while minimizing error.

[0110] Another advantage of using an RLS-type adaptive algorithm to estimate insulation resistance is that it allows for optimized computational costs because it does not require matrix inversion. Therefore, it allows for the integration of this estimation function without increasing the computational power in the BMS, and thus without incurring any additional cost for this still essential function.

[0111] To avoid instability, activation of the insulation resistance R is only performed when the following conditions are met. i Update of the estimated value:

[0112] |ΔU m (k)|>ΔU m_min

[0113] This condition allows the algorithm to be robust to measurement noise and ensures that the voltage U d and U m Synchronization;

[0114] ΔU m (k).ΔU m (k-1)≥0

[0115] This condition allows the algorithm to be robust to sudden transitions.

[0116] |ΔU bat (k)|<ΔU bat_max

[0117] This condition makes the algorithm robust to changes in total battery voltage.

Claims

1. A method for determining the insulation resistance (30) between a point (3) on a high-voltage battery (1) of an electric or hybrid vehicle and the vehicle body (2), the method comprising the steps of: A controllable DC voltage source (41) is provided, connected to a single first terminal (HV-) of the vehicle body and the battery. A first resistor (42) connected in series with the voltage source is provided between the single first terminal and the vehicle body. This first resistor is capable of limiting the current flowing into the single first terminal of the battery. A second resistor (44) is provided to be connected in series between the first resistor and the voltage source. Provide a measuring device capable of measuring the voltage across the second resistor. Different continuous voltage setpoint values ​​are applied between the vehicle body and a single first terminal of the battery by the voltage source. The measuring device acquires a voltage measurement signal, which represents the measured voltage across the second resistor for each continuously applied voltage setpoint value. The value of the insulation resistance is calculated based on the voltage measurement signal. The method is characterized in that it implements an adaptive filtering step for the voltage measurement signal, and the method includes the following steps: recursively estimating the vector of transfer function coefficients of the adaptive filter, providing an update of the filter coefficients, and calculating the value of the insulation resistance based on the estimated value of the vector of transfer function coefficients of the adaptive filter.

2. The method as described in claim 1, characterized in that, The value of the insulation resistance is calculated using the following formula: Among them, R i This is the value of the insulation resistance, R. d This is the value of the first resistor, R. m Is this the second one? The value of the resistor, and It is an estimate of the vector of transfer function coefficients of the adaptive filter.

3. The method as described in claim 2, characterized in that, The vector of transfer function coefficients of the adaptive filter is estimated by considering the transfer functions of the input signal u(k) and the output signal y(k), which are defined as follows: u(k)=U d2 (to d1 (k)=ΔU d (k) y(k)=U m2 (to m1 (k)=ΔU m (k) Among them, U m1 (k) and U m2 (k) respectively correspond to these voltage setpoint values ​​U applied continuously by the voltage source. d1 (k) and U d2 (k) is the voltage value measured across the two ends of the second resistor.

4. The method as described in claim 3, characterized in that, The input signal to the filter transfer function is replaced with the input signal defined as follows: Among them, U bat1 and U bat2 These correspond to the total battery voltage values ​​considered for each voltage setpoint value continuously applied by the voltage source, in order to account for the change in total battery voltage during each iteration of the applied voltage setpoint value.

5. The method as described in claim 1, characterized in that, The method includes the step of calculating the position of the insulation resistance relative to a single first terminal of the battery based on the estimated value.

6. The method as described in claim 5, characterized in that, The vector of transfer function coefficients of the adaptive filter is estimated by considering the transfer functions of the input signal u(k) and the output signal y(k), which are defined as follows: u(k)=U bat ·(IN m2 -IN m1 ) y(k)=U m1 U d2 -U d1 U m2 Among them, U m1 and U m2 These correspond to the voltage setpoint values ​​U that are continuously applied by the voltage source. d1 and U d2 The voltage value measured across the second resistor, and U bat This is the total battery voltage.

7. The method as described in claim 6, characterized in that, The input signal to the filter transfer function is replaced with the input signal defined as follows: u(k)=U m2 ·IN bat1 -IN m1 ·IN bat2 Among them, U bat1 and U bat2 These correspond to the total battery voltage values ​​considered for each voltage setpoint value continuously applied by the voltage source, in order to account for the change in total battery voltage during each iteration of the applied voltage setpoint value.

8. The method as described in any one of the preceding claims, characterized in that, The filtering step is performed using a recursive least squares algorithm.

9. The method according to any one of claims 1-7, characterized in that, The method includes using a single filter setting parameter, which represents the forgetting factor of the filter and consists of real coefficients with values ​​between 0 and 1.

10. An apparatus for determining the insulation resistance (30) between a point (3) on a high-voltage battery (1) of an electric or hybrid vehicle and the vehicle body (2), the apparatus comprising a detection circuit (4) for detecting an insulation fault between the battery and the vehicle body, the detection circuit comprising: A controllable DC voltage source (41) capable of applying a voltage setpoint (U) between a single first terminal (HV-) of the battery and the vehicle body. d ); a first resistor (42) connected in series with the voltage source between the single first terminal and the vehicle body, the first resistor being capable of limiting the current flowing into the single first terminal of the battery; a second resistor (44) connected in series between the first resistor and the voltage source; and a measuring device capable of measuring the voltage (U) across the second resistor. m The device includes a control unit (46) capable of applying different continuous voltage setpoint values ​​between the vehicle body and a single first terminal of the battery via the voltage source; capable of acquiring a voltage measurement signal via the measuring device, the voltage measurement signal representing the measured value of the voltage across the second resistor for each continuously applied voltage setpoint value; and capable of calculating the value of the insulation resistance based on the voltage measurement signal. The control unit is characterized by including an adaptive filtering module for implementing adaptive filtering processing according to the method of any one of the preceding claims.

Images