INSULATION MONITORING IN A HIGH-VOLTAGE SYSTEM, ESPECIALLY IN A MOTOR VEHICLE

Description

[0001] The present invention relates to insulation monitoring in a high-voltage system, in particular in a motor vehicle.

[0002] Modern commercial vehicles increasingly use high-voltage systems that contain multiple high-voltage components. It is important that the high-voltage components have sufficient insulation resistance relative to ground. Therefore, it is common practice to measure the insulation resistance of the high-voltage components using an insulation tester, with a monitoring unit monitoring the measured insulation resistance. However, until now, the insulation resistance value has only been measured at relatively long intervals.

[0003] All insulation begins to deteriorate with age, once first used. Insulation in any given application is designed to provide excellent service for many years under normal operating conditions. However, abnormal working conditions or, for example, vehicle accidents have a damaging effect that, if left unchecked, can accelerate deterioration, eventually causing insulation failure or even leading to dangerous conditions in vehicles.

[0004] In electromechanical battery systems, especially in electrically powered vehicles, electrical loads are switched using large electrical switches (relays, contactors). In dangerous situations that can arise due to accidents or faulty operating conditions, the battery systems and the systems connected to them outside the battery systems (cables or loads) must be protected, especially against a fire that could then spread throughout the vehicle. This protection can be ensured using known shutdown systems, but this requires timely information about the high-voltage insulation.

[0005] High-voltage protection is required. This is ensured by high-voltage insulation from the surrounding ground. Decreasing or missing high-voltage insulation from the ground can lead to hazardous environments or situations, especially for people. A possible insulation fault can have various causes, such as the manufacturing of individual components, the assembly of components, the interconnection of components, the aforementioned aging, and incorrect operation.

[0006] In automobiles, the ground is designated "terminal 31" across all manufacturers (terminal designation). An insulation fault causes a fault current from the high-voltage positive to the high-voltage negative via terminal 31, thereby generating voltage drops in the system that can be dangerous. This creates life-threatening fault sources.

[0007] There is the high-voltage battery as a self-contained component, and the high-voltage battery with a connected circuit (via galvanic isolators in high-voltage positive and negative) with the electrical consumers. The high-voltage battery itself normally has no capacitance to the vehicle's ground. The high-voltage battery with connected consumers has a parasitic capacitance from ground to the high-voltage positive and negative.

[0008] Insulation measurements must produce reliable results with very small capacitances (a few nF) and large capacitances (a few µF), which are a measure of the insulation. Insulation is considered faulty if it does not adequately prevent current flow in undesired paths. This includes current flow across the inner and outer surfaces of the insulation (leakage current) and many other causes. Pinholes and cracks, for example, can develop in the insulation, or moisture and other contaminants can penetrate the surface(s). These contaminants easily ionize under an applied voltage, creating a pathway for the low resistance of leakage current, which can vary depending on the level of contamination and ambient humidity compared to dry, uncontaminated surfaces.

[0009] Various currents are known in the art, the terminology of which is briefly explained below. The capacitive charging current is the current required to charge the capacitance of the insulation under test. The current is initially large but relatively short-lived, falling close to zero depending on the charge of the object under test. An insulating material is charged in the same way as a non-conductor in a capacitor. An absorption current consists of up to three currents that fall close to zero at a decreasing rate over several minutes.

[0010] A surface leakage current occurs when the insulation surface is contaminated, for example, with moisture or salts. The current is constant over time and depends on the degree of contamination present, which in turn depends on the temperature. However, this is often ignored as a single current, including the conduction current, among the total leakage current. However, it is the ISO current that is relevant.

[0011] The insulation current is continuous through the insulation under consideration and is usually represented by a very high resistance parallel to the insulation capacitance. It is a component of the leakage current, which is the current measured when the insulation is fully charged and complete absorption has occurred.

[0012] The insulation resistance measurement for electrical systems according to Voltage Class B must also be carried out during vehicle conditioning under the conditions under which the lowest value is expected. This is done using an insulation resistance tester in accordance with DIN EN 61557-2 VDE 0413-2:2008-02 (Electrical safety in low-voltage systems up to 1,000 V AC and 1,500 V DC - Equipment for testing, measuring, or monitoring protective measures, Part 2) by measuring the insulation resistance at a measurement voltage that corresponds to the system's HV voltage, so that, for example, the resistance value R ISO 500 is determined at 500 V.

[0013] Another disadvantage of insulation measurement is that a current measurement is required at one of the battery terminals, for example, to compensate for driving disturbances. Furthermore, the current and voltage measurements at the battery must be synchronized. The battery's resistance, which depends on the operating point, must be determined from these values. The determined values ​​can then be used to mathematically compensate for the voltage fluctuations. Furthermore, knowledge of the capacitances is required.

[0014] The small and large capacitances in the system also influence the duration of the insulation evaluation. Insulation measurements are typically performed in excessively long time windows of up to 30 seconds or more. In automotive engineering, insulation evaluation must be available within a few seconds. The conventional and well-known measurement methods with long measurement times are not effective.

[0015] Furthermore, there is a need to be able to determine insulation without specifying capacitance values, so that precise knowledge of the capacitance, regardless of whether it is large or small, is not necessary. Furthermore, disturbances such as driving disturbances on the measurement signal, charging, or discharging processes should not negatively influence the measurement.

[0016] Various methods for determining the insulation strength as well as aspects or devices relating to such methods are known in the prior art, for example from the documents US 2019 / 242932 A1, DE 10 2019 214532 A1 and US 2012 / 119754 A1.

[0017] The invention is therefore based on the object of making known high-voltage systems correspondingly safe. Furthermore, the invention is based on the object of providing a corresponding monitoring method that enables rapid determination of the insulation condition in DC high-voltage systems in vehicles without knowing the vehicle's capacity.

[0018] This object is achieved by the method according to the main claim and the insulation measuring device according to the independent claim 7.

[0019] A basic idea of ​​the present invention can be divided into the following aspects: An insulation determination with increased accuracy and shortened determination time, combined with a capacitance calculation in a three-stage concept: 1) Compensation of driving disturbances 2) Final value estimation of an e-function as a preliminary calculation 3) Iterative optimization of the insulation resistance to be determined to the value with the smallest square error

[0020] A further aspect of the present invention relating to the final value estimation of the insulation resistance consists in assuming a value (insulation value) that can only be determined using static methods after a certain period of time (measurement time, e.g., 30 s) on the basis of measured values ​​determined at short intervals (e.g., every 500 ms) as the support points of an e-function, determining the e-function and its limit value therefrom, and determining the insulation value (insulation resistance for either HV plus to ground or HV minus to ground) of the high-voltage system from the limit value of the e-function. Due to the respective characteristic curve of the corresponding e-function, the limit value can be extrapolated from an e-function and therefore a final value estimation can be carried out as a preliminary calculation.

[0021] The further advantages to be achieved with the invention can be stated as follows: The determined resistance insulation value has a low tolerance and therefore a high quality of result. Determination of the capacities is possible. Insulation value determination is independent of the capacity value specifications installed in the system (vehicle). The insulation value determination reacts to changing capacity values ​​(capacity value changes due to aging processes in the vehicle, structural changes, for example to the wiring harness, installation of additional devices). The determined final value informs the system as to whether the insulation quality is sufficient and whether the drive can be started and used again during operation.

[0022] The present invention proposes the following procedure for the three-stage process: 1) Capacity calculation

[0023] The following metrics should be used for this purpose: Record measured values ​​of the battery voltage between the HV positive and HV negative terminals of the battery. Record measured values ​​for the voltage from vehicle ground to battery potential (HV positive terminal) or HV negative with reference to terminal 31. Filter the measured battery voltage value so that the frequency response, depending on the driving condition, charging / discharging, and load, corresponds to the voltage from vehicle ground to the battery terminal (HV positive terminal). In other words, this means obtaining the value at which the voltage between the HV positive and HV negative terminals corresponds to the voltage between ground and the HV positive terminal.

[0024] Thus, according to the idea of ​​the invention, the following measured values ​​are to be recorded: the unfiltered battery voltage measured value, the filtered battery voltage measured value with the same behavior or frequency behavior as the battery voltage between vehicle ground to the battery pole HV-Plus and the voltage between vehicle ground to the battery pole HV-Minus.

[0025] On this basis, a calculation is carried out according to a suitable calculation algorithm, which can preferably look like this: Record the voltage from vehicle ground to battery potential (HV positive terminal) Calculate the value Vx_comp[n] as follows: Vx _ comp n = 1 2 Vbatmean n − Vbat _ dc n + Vx _ dc n where Vx_comp[n]: represents the compensated measurement voltage, in which the voltages caused by driving disturbances have already been calculated Vbatmean[m]: filtered battery voltage measurement value with frequency response equal to the battery voltage vehicle ground to the battery terminal Vbat_dc[n]: unfiltered battery voltage measurement value Vx_dc[n]: unfiltered voltage vehicle ground to the battery terminal 2) Final value estimation of an e-function as a preliminary calculation

[0026] Another aspect of the invention is to use the e-function to determine the final values ​​in advance using its properties.

[0027] The e-function can be determined, for example, using suitable support points. The final value estimation supports the iterative calculation of the final value in a shorter time. 3) Iterative optimization of the insulation resistance to be determined

[0028] In this step, the optimized final value of the e-function is iteratively approximated in order to determine the insulation resistance as accurately as possible. The final value estimate serves as the starting point for the subsequent final value calculation and has the advantage that the final value calculation begins at a reduced level of squared error.

[0029] The following equation can be used to advantage: Y t = A e t tau + EE where the following sizes are considered: Y:Measured values ​​A:Parameter Tau:Parameter time constant EE:End value e-function t:Measurement time.

[0030] A precalculated pseudo-inverse is used. A starting value for EE is used for the iterative calculation. This is usually a relatively large initial value for the insulation resistance.

[0031] The features disclosed above can be combined as desired, as long as this is technically possible in the implementation of the invention and they do not contradict each other technically, even if this combination of features is not expressly disclosed.

[0032] Other advantageous developments of the invention are characterized in the subclaims or are presented in more detail below together with the description of the preferred embodiment of the invention with reference to the figures. They show: Fig. 1 a schematic structural embodiment of the invention Fig. 2 a flow chart showing the process steps for determining the insulation strength and Fig. 3 schematic curves 1 to 4 which were used to explain the iterative calculation.

[0033] The figures and curves are merely exemplary schematic. Like reference numerals in the figures indicate like functional and / or structural features. The invention will be described below with reference to the Fig. 1 to 3 explained in more detail using exemplary embodiments.

[0034] The Fig. 2 shows a flow chart showing the three-stage concept for determining insulation strength.

[0035] In the Figure 1A purely schematic structural embodiment of the invention is shown using an exemplary DC high-voltage system, in particular in a motor vehicle, with a plurality of high-voltage components 10, 11, 12, wherein a high-voltage energy storage device (not shown in detail) is provided, which is connected to the high-voltage components 10, 11, 12 via an electrical line, as well as an insulation measuring device 20 for measuring the voltage between HV plus or HV minus of the high-voltage components 10, 11, 12 and the vehicle ground, and at least one evaluation unit 30 for determining the insulation resistance based on an iterative calculation, as explained in the following exemplary algorithm for the iterative optimization of the insulation resistance to be determined. However, the invention is not limited to precisely this solution; other iterative approaches are also conceivable.

[0036] The following equation can be used to advantage: Y t = A e t tau + EE where the following sizes are considered: Y:Measured values ​​A:Parameter Tau:Parameter time constant EE:End value e-function t:Measurement time.

[0037] A precalculated pseudo-inverse is used. A starting value for EE is used for the iterative calculation. This is usually a relatively large initial value for the insulation resistance. Step 1:

[0038] In a first step, the e-function is linearized by taking the logarithm, as follows: ln Y t − EE = ln A e − t tau ln Y t − EE = ln A + t / tau

[0039] As a result, the solution system now has the form of a linear equation of the general form: B = A ∗ x ln Y t 0 − EE ⋮ ln Y tn − EE = 1 t 0 ⋮ ⋮ 1 tn ∗ ln A τ Step 2:

[0040] In a subsequent step, the matrix multiplication with the pseudoinverse is performed, the result being tau and In(A). Step 3:

[0041] In a further subsequent step, the squared error is calculated (left side and right side of matrix 1). Step 4:

[0042] In a further step, the iteration of EE takes place: Step 5:

[0043] In this step, a comparison is made as to whether the curve can be approximated even better as a straight line (for very small time constants, the system has step-like behavior, whereby a straight line solution is the better approach). Iteration with EE = 0 Tau = 0 Exp (In(A)) is the solution for the line Calculate the square error of the line solution Step 6:

[0044] The iteratively determined value for EE with the smallest squared error is used for the subsequent calculation. The parameters EE and tau are used for the resistance calculation (EE) and for the capacitance calculation (tau). Tau determines the system's time response and represents the filter constant.

[0045] The time behavior of the system in Laplace is as follows: (Example for the voltage to the negative battery terminal) U mass _ bat U bat = R neg R neg + R pos ∗ 1 + s ∗ R pos ∗ C pos 1 + s ∗ R neg ∗ R pos R neg + R pos ∗ C pos + C neg U mass_bat = Voltage of battery mass to negative battery terminal U asked = Battery voltage R neg = negative insulation resistance R pos = positive insulation resistance C neg = negative capacity C pos = positive capacity

[0046] An iterative calculation is performed to find the optimal combination of values ​​for EE, C, A, and tau (according to the principle of least squared error). EE is changed with each iteration. Tau or In(A) could also be iterated. Using the pseudoinverse, the parameter pair A and tau with the smallest squared error is calculated for a given EE.

[0047] A separate tau is calculated for each of the rising and falling curves (tau1 and tau2). To adjust the time constant of the mean filter, both calculations of tau1 and tau2 are combined. In the first step, a mean value is used. Curves 3 and 4 show the squared errors.

[0048] With the final value estimation, the iterative final value calculation is started close to the optimum, i.e. with minimal square error, whereby the aim of the invention is possible with a minimized time requirement of the iteration, thus the calculation of the insulation resistance.

[0049] The insulation resistance is calculated using the results of the iterative final calculations (curves 1 and 2, and curves 3 and 4, respectively, rising and falling curves). Using the tau and the calculated insulation resistance, the capacitances can also be calculated. Curves 1, 2, 3, and 4 are examples; in each curve, a theoretical curve is identical to the calculated curve shown with the optimal EE parameter. Curves 3 and 4 show the root mean square error for the respective EE used. The Y-axis shows the voltage from the vehicle ground to the battery terminal divided by the battery voltage.

[0050] The resistance calculation can be done with one or two curves.

[0051] The insulation measuring device 20 from the Figure 1 is therefore after a

[0052] Embodiment of the invention for determining the insulation strength of the electrical insulation of the DC high-voltage components 10, 11, 12 in the DC high-voltage system, in particular for such a high-voltage system in a vehicle having an electrical DC energy storage device, wherein the evaluation unit 30 is specifically designed to process the measured values ​​from the insulation measuring device 20 according to the method described above in order to determine the insulation resistance of the DC high-voltage system therefrom.

[0053] The invention is defined by the appended claims.

Claims

1. A method for ascertaining the insulation strength of the electrical insulation of DC high-voltage components (10, 11, 12) in a DC high-voltage system, in particular for a high-voltage system in a vehicle having an electrical DC energy store, with the following steps: a) ascertaining the influence of the compensation variables of the driving disturbances occurring in the vehicle on the insulation strength to be ascertained; b) performing a final value estimation of the insulation resistance based on an e-function as a preliminary calculation for use as an initial value of a subsequent iterative calculation, and c) iteratively ascertaining the insulation resistance to be determined by means of optimising the least squared error for the insulation resistance.

2. The method of claim 1, wherein the following equation is utilised as the e-function: Y t = Ae t tau + EE wherein the following variables are used Y: sensed measurement values A: parameter Tau: time constant EE: e-function final value t: measuring time.

3. The method of any one of the preceding claims, characterised in that the capacitance of the DC high-voltage system in the vehicle is ascertained by way of calculation.

4. The method of any one of the preceding claims, characterised in that the battery voltage between the HV positive pole and HV negative pole of the DC energy store are sensed as measurement values, as well as measurement values for the voltage from vehicle earth to battery potential (HV positive pole) are acquired, and preferably, the filtering of the measurement value of the battery voltage takes place so that the frequency response of the battery voltage corresponds to the vehicle earth to battery pole (HV positive pole) voltage.

5. The method of the preceding claim, characterised in that the following calculation of the variable Vx_comp[n] is utilised in ascertaining the compensation variables: Vx _ comp n = 1 2 Vbatmean n − Vbat _ dc n + Vx _ dc n wherein Vx_comp[n]: represents the compensated measuring voltage at which the voltages contingent upon driving disturbances have already been calculated out Vbatmean[n]: filtered battery voltage measurement value with frequency response equal to the vehicle earth to battery pole battery voltage Vbat_dc[n]: unfiltered battery voltage measurement value Vx_dc[n]: unfiltered vehicle earth to battery pole voltage.

6. The method of any one of the preceding claims, characterised in that the following steps are performed a. in an initial step, the linearisation of the e-function takes place by logarithmising as follows: ln Y t − EE = ln Ae t tau ln Y t − EE = ln A + t / tau whereby a solution system underlying the calculation now has the form of a linear equation of the general form: B = A ∗ x ln Y t 0 − EE ⋮ ln Y tn − EE = 1 t 0 ⋮ ⋮ 1 tn ∗ ln A τ b. in a subsequent step, the matrix multiplication from step a) takes place with the pseudoinverse, wherein the result is tau (time constant) and the variable In(A); c. in a subsequent step, the calculation of the squared error takes place; d. then the iteration of EE takes place; e. optionally, the comparison takes place as to whether the ascertained curve can be approximated even better as a straight line, and f. then the iteratively ascertained value for EE, which has the respectively least squared error, is used for the further subsequent calculation for ascertaining the insulation strength and optionally for the determination of the capacitance of the system by way of calculation.

7. An insulation measuring device (20) for determining the insulation strength of the electrical insulation of DC high-voltage components (10, 11, 12) in a DC high-voltage system, wherein the insulation measuring device (20) is configured with at least one evaluation unit (30) configured to process the measurement values from the insulation measuring device (20) according to the method of any one of claims 1 to 6 in order to determine the insulation resistance of the DC high-voltage system therefrom.