CN113655281B - Insulation resistance value detection method during disconnection of electric automobile power battery - Google Patents
Insulation resistance value detection method during disconnection of electric automobile power battery Download PDFInfo
- Publication number
- CN113655281B CN113655281B CN202110932123.9A CN202110932123A CN113655281B CN 113655281 B CN113655281 B CN 113655281B CN 202110932123 A CN202110932123 A CN 202110932123A CN 113655281 B CN113655281 B CN 113655281B
- Authority
- CN
- China
- Prior art keywords
- voltage
- model
- value
- parameter
- bus voltage
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R27/00—Arrangements for measuring resistance, reactance, impedance, or electric characteristics derived therefrom
- G01R27/02—Measuring real or complex resistance, reactance, impedance, or other two-pole characteristics derived therefrom, e.g. time constant
- G01R27/025—Measuring very high resistances, e.g. isolation resistances, i.e. megohm-meters
Abstract
The invention discloses an insulation resistance value detection method when an electric automobile power battery is disconnected, which actually considers the influence of a Y capacitor in an electric automobile high-voltage system on a measurement process and establishes a function model of sampling voltage and bus voltage. The two function models are simplified, and parameters of the simplified busbar voltage function model are decoupled, so that the calculated amount is greatly reduced, and the algorithm is more suitable for being used in a vehicle-mounted embedded system. And identifying the stable value of the sampling voltage and the stable value of the bus voltage by adopting a forgetting factor recursive least square algorithm, and finally calculating the insulation resistance values of the positive electrode and the negative electrode according to the identification result. On the basis of a low-frequency signal injection method, the invention establishes the function model of the sampling voltage and the bus voltage, simplifies the function model, performs parameter decoupling on the simplified bus voltage model, applies a forgetting factor recursive least square algorithm in insulation detection, and eliminates the influence of the Y capacitance.
Description
Technical Field
The invention belongs to the technical field of electric automobile power batteries, and particularly relates to an insulation resistance value detection method based on a forgetting factor recursive least square algorithm when the electric automobile power battery is disconnected, and a non-decoupling and decoupling mode without a voltage source.
Background
As automobile pollutant emissions become more and more severe, electric automobiles become more popular. The lithium ion battery has higher energy density and power density, and the high voltage can obviously improve the energy utilization rate. Therefore, most electric vehicles use a high-voltage lithium ion battery as a power battery. Because the high-voltage lithium ion battery belongs to a high-voltage direct current system, the high-voltage insulation performance of the power battery has important significance for safe driving of the electric automobile. The running condition of the electric automobile is complex, and the electric automobile has complex environments such as high temperature, high humidity, high salt fog, vibration collision and the like. These factors can all lead to the insulation performance between the positive and negative buses of the power battery and the electric chassis of the vehicle to be reduced, so that the electric potential of the chassis of the vehicle is increased or reduced, the normal operation of the low-voltage component and the high-voltage component is influenced, and the personal safety of drivers and passengers is endangered. Therefore, the method is of great importance to detection of the insulation resistance of the power battery of the electric automobile, is the core content of the electric safety technology of the electric automobile, and has great significance to safety of the whole automobile and drivers.
Unlike conventional fuel vehicles, electric vehicles are composed of numerous high voltage components, and in order to isolate electromagnetic interference between a high voltage bus and a high voltage device, a Y capacitor is generally added at a power supply inlet of the high voltage device. For an electric automobile, on one hand, the Y capacitance attached to a plurality of high-voltage components on the automobile objectively causes that Y capacitance is necessarily existed between two poles of a high-voltage bus and an automobile chassis; on the other hand, an equivalent Y capacitance is formed between the high-voltage bus cable penetrating through the vehicle body and the chassis. The system Y capacitance bridged between the high-voltage bus and the electric chassis can change the impedance characteristic between the high-voltage bus and the electric chassis, and influence the insulation performance between the high-voltage bus and the ground.
The traditional insulation detection method can be divided into a passive type and an active type according to whether an external power supply exists or not, and the common method mainly comprises a bridge method and a signal injection method. The bridge method belongs to passive type, its basic principle is that a current-limiting resistor is connected between the positive and negative buses of the power battery and the electric chassis of the vehicle, the size of the current-limiting resistor is changed by an electronic switch, the partial pressure of the measured resistor is measured under the condition of connecting different resistors, and finally the insulation resistor is calculated by solving an equation. The bridge method has simple circuit and can directly calculate the insulation resistance, but the method has more defects: an electronic switch added in the circuit can introduce switching noise to the high-voltage bus; can only be used when the power battery is connected; and a Y capacitor exists between the two poles of the power battery and the electric chassis, and the measurement period is increased to improve the insulation resistance detection precision, so that the response speed of the measurement system is reduced. The signal injection method belongs to an active type, and the basic principle is that a detection signal is injected into a high-voltage system through a vehicle electric chassis, and the insulation resistance value is calculated by measuring a voltage signal on a sampling resistor. In contrast to the bridge method, this method does not introduce switching noise on the high voltage bus. At present, an injection method based on low-frequency signals is widely applied to electric automobile insulation detection. However, a Y capacitor exists between the two poles of the power battery and the electric chassis, and due to the charging and discharging processes of the Y capacitor, the sampling voltage cannot reach a stable value due to an excessively short signal period, so that the measurement result of the insulation resistance value is smaller, the system is caused to alarm in a false manner, and the response speed of the measurement system is reduced due to an excessively long signal period. And the Y capacitance value is easily influenced by external factors such as vehicle part replacement, environmental temperature and humidity change and the like, and further influences the insulation resistance measurement precision and the measurement period. In addition, the working condition of the electric automobile is very complex, signals are easy to be interfered by noise, and further measurement accuracy is reduced.
The invention with publication number CN109100618A utilizes a passive balance bridge-unbalance bridge method to measure the insulation resistance value of the positive electrode and the negative electrode, judges the magnitude relation of the voltage of the positive electrode bus and the negative electrode bus to the ground through AD sampling, then accesses a sampling resistor at the side with smaller voltage, and lists equation calculation and judges the insulation performance. The method has the advantages of simple structure and low cost. The disadvantage is that the situation that the insulation resistance values of the positive electrode and the negative electrode are equal is not considered; the influence of the Y capacitance on the voltage signal on the sampling resistor is not considered; the insulation resistance cannot be calculated when the power battery is disconnected.
The invention with the publication number of CN103076497A calculates the insulation resistance of the anode and the cathode by detecting the change of the charge-discharge voltage value of the capacitor, has the advantages that a detection system is effectively isolated from a battery system by adopting a capacitor isolation mode, partial signal interference can be filtered, the detection precision is improved to a certain extent, and the defects that the charge-discharge of the capacitor leads to longer detection period and lower timeliness are overcome.
The invention with the publication number of CN105738701A relates to the process of measuring the insulation resistance after weighting the fluctuation of the total voltage of the battery, and has the advantages of small fluctuation weight and large fluctuation weight. The method has the advantages of reducing the influence of battery voltage fluctuation on measurement and improving the measurement precision. The defects are that: the detection time is long, the calculation complexity is high, the real-time performance of measurement is insufficient, and the influence of the Y capacitance on a measurement system is not considered.
Therefore, when the power battery is disconnected, how to eliminate the influence of the Y capacitor on the measurement process and accurately detect the insulation resistance becomes a problem to be solved by the staff in the same row.
Disclosure of Invention
The invention mainly aims to provide an insulation resistance value detection method when an electric automobile power battery is disconnected, and particularly relates to an insulation resistance value detection method based on a forgetting factor recursive least square algorithm when the electric automobile power battery is disconnected, which can solve the problems, can eliminate the influence of a Y capacitor on a measurement process, and can rapidly and accurately calculate the insulation resistance values of an anode and a cathode.
In order to achieve the above purpose, the technical scheme adopted by the invention is as follows:
the embodiment of the invention provides an insulation resistance value detection method when an electric automobile power battery is disconnected, which comprises the following steps:
s1, constructing an insulation detection circuit model when a power battery with a Y capacitor is disconnected; the insulation detection circuit model includes: an insulation detection equivalent circuit and an electric automobile high-voltage equivalent system;
s2, generating square wave signals by an insulation detection equivalent circuit, injecting the square wave signals into an automobile high-voltage equivalent system through an electric chassis, and passing through a positive-negative insulation resistor R of an electric automobile p 、R n And positive and negative Y capacitance C p 、C n Flow back to the insulation detection, etcAn effect circuit;
s3, establishing a function model of the sampling voltage and the bus voltage according to the reflowed signals;
s4, simplifying the function models of the sampling voltage and the bus voltage respectively, simplifying the model with five parameters into a model with three parameters, and performing parameter decoupling on the simplified bus voltage model;
s5, identifying the sampling voltage by adopting a forgetting factor recursive least square algorithm; identifying the decoupled bus voltage model by adopting a twice forgetting factor recursive least square algorithm;
s6, calculating the insulation resistance of the positive electrode and the negative electrode according to the identification result.
Further, the step S3 includes:
from the signal flowing back, a sampling voltage U is established f The function of the value of (c) at time t is:
establishing bus voltage U pn The function of the value of (c) at time t is:
a 1 、a 2 、a 3 、a 4 、a 5 、b 1 、b 2 、b 3 、b 4 、b 5 is an unknown parameter of the two function models, and e is a constant.
Further, the step S4 includes:
s41, based on the sampling voltage value a when the voltage is required to reach stability when the insulation resistance value is calculated 1 And bus voltage value b 1 And other parameters not related to the calculation of insulation resistance value, the voltage U is sampled f And bus voltage U pn Simplifying the function model of (2);
s42, sampling voltage U f The functional model at time t is simplified as:
U f (t)＝a 1 +a' 2 exp(-t/a' 3 ) (3)
bus voltage U pn The functional model at time t is simplified as:
U pn (t)＝b 1 +b 2 'exp(-t/b 3 ') (4)
wherein a is 1 And b 1 To simplify the steady-state component of the response of the post-model, i.e. a 1 B is the stationary value of the sampled voltage 1 Is a stable value of the bus voltage; a' 2 And b' 2 To simplify the response gain of the post-model, a' 3 And b' 3 Response time constants for the simplified back model; a' 2 、b' 22 、a' 3 And b' 3 The insulation resistance value is not calculated;
s43, sampling voltage U f And (3) converting the continuous time function model of the simplified formula (3) into a discrete time function model, and performing parameter decoupling on the formula (4).
Further, in the step S43, the sampling voltage U is sampled f The continuous time function model of the simplified formula (3) is converted into a discrete time function model, and the method is specifically expressed as follows:
U f (k)＝a 1 (k)+a' 2 (k)exp(-kΔt/a' 3 (k)) (3-1)
wherein Δt is a sampling time interval; sampling voltage U f The discrete time function model of (2) is a nonlinear model, and is linearized by adopting first-order Taylor series expansion and converted into a least square model.
Further, the voltage U is sampled f The discrete time function model of (2) is a nonlinear model, and is linearized by adopting first-order Taylor series expansion and converted into a least square model; comprising the following steps:
u is set to f (k) Expanding the parameter estimation value around the last moment into a Taylor series, and omitting secondary and above items;
sampling voltage U f The simplified function model is converted into a least square model, and the expression is as follows:
U f (k)＝H a (k)A(k)+Y a (k)+e a (k) (3-2)
wherein H is a (k) An observation matrix at the moment k; a (k) is a parameter value at time k, and a (k) = [ a ] 1 (k),a' 2 (k),a' 3 (k)]，Y a (k) E is a constant error term a (k) Is a one-dimensional random observation noise and is a zero-mean value independent Gaussian white noise sequence.
Further, in the step S43, the parameter decoupling of the formula (4) includes:
according to the bus voltage U pn The simplified function model at time t yields U pn At t k Equation of time:
U pn (t k )＝b 1 +b' 2 exp(-t k /b' 3 ) (4-1)
then, the process is carried out,
U pn (t k )＝b 1 +U pn (t k-1 )exp(-Δt/b' 3 )-b 1 exp(-Δt/b' 3 ) (4-2)
wherein Δt is the sampling interval time; u of adjacent two moments pn The expression of (2) is subtracted to obtain the following expression:
ΔU pn (t k )＝exp(-Δt/b' 3 )*ΔU pn (t k-1 ) (4-3)
wherein DeltaU pn (t k ) At t k U at moment of time pn Subtracting t k-1 U at moment of time pn ，ΔU pn (t k-1 ) At t k-1 U at moment of time pn Subtracting t k-2 U at moment of time pn ；
Bus voltage U pn Simplified function model parameters b 1 And parameter b 3 The' decoupling is obtained, and the two least squares models after decoupling are respectively as follows: the first least squares model of the bus voltage is:
ΔU pn (k)＝H b1 (k)*β(k)+e b1 (k) (4-4)
wherein H is b1 (k)＝U pn (t k-1 )-U pn (t k-2 )，β(k)＝exp(-Δt/b' 3 (k))，ΔU pn (k)＝U pn (t k )-U pn (t k-1 )，e b1 (k) A Gaussian white noise sequence which is one-dimensional random observation noise and has zero mean value and is independent;
the second least squares model of the bus voltage is:
U pn (k)＝H b2 (k)*b 1 (k)+Y b (k)+e b2 (k) (4-5)
wherein U is pn (k) Representing t k U at moment of time pn ， E is the state estimation value at time k in the first least square model of the bus voltage b2 (k) Is a one-dimensional random observation noise and is a zero-mean value independent Gaussian white noise sequence.
Further, in the step S5, the identifying the sampling voltage model by adopting a forgetting factor recursive least square algorithm includes:
s51, setting a least square parameter initial value, an error covariance initial value and a forgetting factor parameter of the sampling voltage; the least square parameter initial value and the error covariance initial value are used for the first iteration of the recursive algorithm, and the forgetting factor parameter is used for calculating the forgetting factor of each iteration;
s52, parameter identification, including calculation of forgetting factors, calculation of observation matrixes, calculation of gain matrixes, input of sampling voltage, error covariance matrixes and parameter estimation values a 1 Is updated according to the update of (a);
s53, calculating the parameter estimation value a in the sliding window 1 Mean and standard deviation of (a); when the standard deviation is smaller than a first preset threshold value, the parameter estimation value a 1 As a smoothed value of the sampled voltage.
Further, in the step S5, the identifying the decoupled bus voltage model by adopting a two-time forgetting factor recursive least square algorithm includes:
identifying a first least square model of the busbar voltage to obtain an estimated value of the parameter beta; substituting the identified parameter beta into a second least square model of the bus voltage, and continuously identifying the second least square model to obtain a parameter b 1 Is a function of the estimated value of (2);
the parameter b is subjected to a two-time forgetting factor recursive least square algorithm 1 Identifying, including:
s501, respectively setting a first least square model, a parameter initial value, an error covariance initial value and a forgetting factor parameter of a bus voltage; the least square parameter initial value and the error covariance initial value are used for the first iteration of the recursive algorithm, and the forgetting factor parameter is used for calculating the forgetting factor of each iteration;
s502, identifying a first least square model of the busbar voltage to obtain a parameter estimation value beta; the parameter estimation value beta is input in the next step;
s503, identifying a second least square model of the bus voltage to obtain a parameter estimated value b 1 ；
S504, calculating the parameter estimation value b in the sliding window 1 Mean and standard deviation of (a); when the standard deviation is smaller than a second preset threshold value, the parameter estimation value b 1 As a plateau value for the bus voltage.
Further, in the step S6, the insulation resistance of the positive electrode and the negative electrode is calculated according to the identification result, including:
the insulation resistance value of the positive electrode of the power battery relative to the automobile electric chassis is calculated, and the calculation formula is as follows:
R p ＝R(2R f (U s -a 1 )/(a 1 R+b 1 R f )-1)
the insulation resistance value of the power battery cathode relative to the automobile electric chassis is calculated, and the calculation formula is as follows:
R n ＝R(2R f (U s -a 1 )/(a 1 R-b 1 R f )-1)
wherein R is p The positive insulation resistance value of the electric automobile; r is R n The negative electrode insulation resistance value of the electric automobile; r is the resistance value of a current-limiting resistor in the insulation detection equivalent circuit; r is R f Is a sampling resistor; u (U) s Is the voltage value generated by the pulse signal generator in the positive and negative half periods.
Compared with the prior art, the invention has the following beneficial effects:
the insulation resistance detection method for the disconnection of the power battery of the electric automobile actually considers the influence of the Y capacitor in the high-voltage system of the electric automobile on the measurement process, and establishes a function model of sampling voltage and bus voltage. The function model is simplified, the model with five parameters is simplified into a model with three parameters, and the simplified bus voltage function model is subjected to parameter decoupling, so that the calculated amount is greatly reduced, and the algorithm is more suitable for being used in a vehicle-mounted embedded system. And identifying the model by adopting a forgetting factor recursive least square algorithm, and finally calculating the insulation resistance value of the positive electrode and the negative electrode according to an identification result. On the basis of a low-frequency signal injection method, the invention establishes the function model of the sampling voltage and the bus voltage, simplifies the function model, performs parameter decoupling on the simplified bus voltage model, applies a forgetting factor recursive least square algorithm in insulation detection, and eliminates the influence of the Y capacitance.
Drawings
Fig. 1 is a flowchart of an insulation resistance value detection method when an electric vehicle power battery is disconnected, provided by an embodiment of the invention;
fig. 2 is a schematic diagram of an insulation detection circuit model when a power battery with a Y capacitor provided in an embodiment of the present invention is disconnected;
fig. 3 is a flowchart of an implementation of an insulation resistance detection algorithm according to an embodiment of the present invention.
Detailed Description
The invention is further described in connection with the following detailed description, in order to make the technical means, the creation characteristics, the achievement of the purpose and the effect of the invention easy to understand.
In the description of the present invention, it should be noted that the directions or positional relationships indicated by the terms "upper", "lower", "inner", "outer", "front", "rear", "both ends", "one end", "the other end", etc. are based on the directions or positional relationships shown in the drawings, are merely for convenience of describing the present invention and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a specific direction, be configured and operated in the specific direction, and thus should not be construed as limiting the present invention. Furthermore, the terms "first," "second," and the like, are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.
In the description of the present invention, it should be noted that, unless explicitly specified and limited otherwise, the terms "mounted," "provided," "connected," and the like are to be construed broadly, and may be fixedly connected, detachably connected, or integrally connected, for example; can be mechanically or electrically connected; can be directly connected or indirectly connected through an intermediate medium, and can be communication between two elements. The specific meaning of the above terms in the present invention will be understood in specific cases by those of ordinary skill in the art.
The invention provides a method for detecting insulation resistance value when an electric automobile power battery is disconnected, referring to fig. 1, comprising the following steps:
s1, constructing an insulation detection circuit model when a power battery with a Y capacitor is disconnected; the insulation detection circuit model includes: an insulation detection equivalent circuit and an electric automobile high-voltage equivalent system;
s2, generating square wave signals by an insulation detection equivalent circuit, injecting the square wave signals into an automobile high-voltage equivalent system through an electric chassis, and passing through a positive-negative insulation resistor R of an electric automobile p 、R n And positive and negative Y capacitance C p 、C n Flowing back to the insulation detection equivalent circuit;
s3, establishing a function model of the sampling voltage and the bus voltage according to the reflowed signals;
s4, simplifying the function models of the sampling voltage and the bus voltage respectively, simplifying the model with five parameters into a model with three parameters, and performing parameter decoupling on the simplified bus voltage model;
s5, identifying the sampling voltage model by adopting a forgetting factor recursive least square algorithm; identifying the decoupled bus voltage model by adopting a twice forgetting factor recursive least square algorithm;
s6, calculating the insulation resistance of the positive electrode and the negative electrode according to the identification result.
In the embodiment, the method actually considers the influence of the Y capacitance in the high-voltage system of the electric automobile on the measurement process, and establishes a function model of the sampling voltage and the bus voltage. The two function models are simplified, and parameters of the simplified busbar voltage function model are decoupled, so that the calculated amount is greatly reduced, and the algorithm is more suitable for being used in a vehicle-mounted embedded system. And identifying the stable value of the sampling voltage and the stable value of the bus voltage by adopting a forgetting factor recursive least square algorithm, and finally calculating the insulation resistance values of the positive electrode and the negative electrode according to the identification result. On the basis of a low-frequency signal injection method, the invention establishes the function model of the sampling voltage and the bus voltage, simplifies the function model, performs parameter decoupling on the simplified bus voltage model, applies a forgetting factor recursive least square algorithm in insulation detection, and eliminates the influence of the Y capacitance.
The above steps are described in detail below.
In the steps S1-S2, considering the influence of the Y capacitance between the two poles of the power battery and the vehicle chassis on the insulation resistance measurement, an insulation detection circuit model of the power battery with the Y capacitance when disconnected is constructed on the basis of the detection principle of the low-frequency injection method, as shown in fig. 2. Wherein, the right side is an equivalent schematic diagram of an insulation detection circuit, U s The pulse signal generator is responsible for transmitting square wave signals, and the voltage values generated by the pulse signal generator in positive and negative half periods are U respectively s+ And U s- ，R 1 、R 2 Is a current limiting resistor, and R 1 ＝R 2 ＝R，R f Is a sampling resistor. The left side is the equivalent origin of the high-voltage system of the electric automobileThe management diagram, U is the voltage of the power battery, the disconnection of the switch indicates that the power battery does not supply power to the electric automobile, R p 、R n Respectively the insulation resistance of the positive electrode and the negative electrode of the power battery, C p 、C n Y capacitors with positive and negative poles respectively, U pn The bus voltage is the voltage difference between the positive bus and the negative bus of the electric automobile. The specific detection principle is as follows: the pulse signal generator generates a pulse signal with the amplitude of U s Is injected into a high-voltage system of an automobile through an electric chassis and passes through a positive-negative insulation resistor R of the electric automobile p 、R n And positive and negative Y capacitance C p 、C n Flows back to the detection circuit, and the signal passes through the current limiting resistor R on the sampling circuit 1 And R is 2 Back to the sampling resistor R f . Due to the existence of the system Y capacitor, the sampling voltage U is caused f And bus voltage U pn And no longer a square wave signal.
In the above step S3, the voltage U is sampled f The function of the value of (c) at time t is:
bus voltage U pn The function of the value of (c) at time t is:
a 1 、a 2 、a 3 、a 4 、a 5 、b 1 、b 2 、b 3 、b 4 、b 5 is an unknown parameter of the two function models, and e is a constant. And identifying the model parameters by adopting a variable forgetting factor recursive least square algorithm. Considering that the two models respectively have five parameters, if the two models are directly identified, a five-dimensional least square algorithm needs to be established, the calculated amount of the algorithm is large, the complexity is high, and the algorithm is inconvenient to use on a vehicle-mounted embedded system. In addition, as the insulation resistance value is calculated, the sampling voltage value and the bus voltage value when the voltage reaches a stable state are needed, namely the voltage is equal to the voltage of the busA in the individual function model 1 And b 1 And the other parameters do not involve the calculation of the insulation resistance value. The identification process can be simplified to a process in which only the second half of the voltage tends to be stationary regardless of the former portion of the voltage waveform, and the waveform that tends to be stationary is in the form of an exponential function.
In the above step S4, the voltage U is sampled by this idea f And bus voltage U pn The above-described complete function model of (c) is simplified.
Sampling voltage U f The functional model at time t is simplified as:
U f (t)＝a 1 +a' 2 exp(-t/a' 3 ) (3)
bus voltage U pn The functional model at time t is simplified as:
U pn (t)＝b 1 +b' 2 exp(-t/b' 3 ) (4)
wherein a is 1 And b 1 To simplify the steady-state component of the response of the post-model, i.e. a 1 B is the stationary value of the sampled voltage 1 Is a plateau value of the bus voltage. a' 2 And b' 2 To simplify the response gain of the post-model, a' 3 And b' 3 To simplify the response time constant of the post-model, the response time constant is determined by a' 2 、b' 2 、a' 3 And b' 3 Not taking part in the calculation of the insulation resistance value, so they are not considered and only need to be identified to obtain a 1 And b 1 And (3) obtaining the product.
Wherein, the simplified process for equation (3) is as follows:
will sample the voltage U f The simplified continuous time function model is converted into a discrete time function model, and the expression is as follows:
U f (k)＝a 1 (k)+a' 2 (k)exp(-kΔt/a' 3 (k)) (3-1)
where Δt is the sampling time interval. Sampling voltage U f The discrete time function model of (2) is a nonlinear model, and is linearized by adopting a first-order Taylor series expansion in consideration of computational complexity and estimation accuracy and converted into a least square model。
The specific method is as follows: u is set to f (k) The parameter estimation values around the last moment are expanded into a Taylor series, and secondary and above items are omitted.
Sampling voltage U f The simplified function model is converted into a least square model, and the expression is as follows:
U f (k)＝H a (k)A(k)+Y a (k)+e a (k) (3-2)
wherein H is a (k) An observation matrix at time k, a (k) is a parameter value at time k, and a (k) = [ a ] 1 (k),a' 2 (k),a' 3 (k)]，Y a (k) E is a constant error term a (k) Is a one-dimensional random observation noise and is a zero-mean value independent Gaussian white noise sequence.
The parameter decoupling process for the above equation (4) is as follows:
according to the bus voltage U pn The simplified function model at time t can yield U pn At t k Equation of time:
U pn (t k )＝b 1 +b' 2 exp(-t k /b' 3 ) (4-1)
then, the process is carried out,
U pn (t k )＝b 1 +U pn (t k-1 )exp(-Δt/b' 3 )-b 1 exp(-Δt/b' 3 ) (4-2)
where Δt is the sampling interval time. U of adjacent two moments pn The expression of (2) is subtracted to obtain the following expression:
ΔU pn (t k )＝exp(-Δt/b' 3 )*ΔU pn (t k-1 ) (4-3)
wherein DeltaU pn (t k ) At t k U at moment of time pn Subtracting t k-1 U at moment of time pn ，ΔU pn (t k-1 ) At t k-1 U at moment of time pn Subtracting t k-2 U at moment of time pn 。
Through the above process, the bus voltage U pn Simplified function model parameters b 1 And parameter b 3 'decoupling'. The two least squares models after decoupling are each as follows. The first least squares model of the bus voltage is:
ΔU pn (k)＝H b1 (k)*β(k)+e b1 (k) (4-4)
wherein H is b1 (k)＝U pn (t k-1 )-U pn (t k-2 )，β(k)＝exp(-Δt/b' 3 (k))，ΔU pn (k)＝U pn (t k )-U pn (t k-1 )，e b1 (k) Is a one-dimensional random observation noise and is a zero-mean value independent Gaussian white noise sequence.
The second least squares model of the bus voltage is:
U pn (k)＝H b2 (k)*b 1 (k)+Y b (k)+e b2 (k) (4-5)
wherein U is pn (k) Representing t k U at moment of time pn ， E is the state estimation value at time k in the first least square model of the bus voltage b2 (k) Is a one-dimensional random observation noise and is a zero-mean value independent Gaussian white noise sequence.
Model parameter b of simplified bus voltage function 1 And parameter b' 3 After decoupling, firstly identifying a first least square model of the bus voltage to obtain an estimated value of the parameter beta, then substituting the identified parameter beta into a second least square model of the bus voltage, and continuously identifying the second least square model to obtain the parameter b 1 The estimated value of (a), namely, the parameter b is subjected to a two-time variable forgetting factor recursive least square algorithm 1 And (5) identifying.
Stationary value a of sampled voltage 1 And a plateau value b of the bus voltage 1 And (5) identifying. The identification of the sampled voltage plateau and the bus voltage plateau are when the algorithm is runningSimultaneously. And firstly setting parameters w and epsilon for judging algorithm convergence when the algorithm starts, and judging whether the algorithm reaches convergence after each iteration is ended or not, and stopping iteration if the algorithm reaches convergence. The identification of the sampled voltage plateau and the identification of the bus voltage plateau use the same parameters converged by the judgment algorithm, namely the same sliding window w and the same convergence accuracy epsilon.
In the step S5 described above: sample voltage plateau value a 1 The identification of (2) includes the following six steps.
The first step: initializing. Setting the initial value of least square parameter of sampling voltageInitial value of error covariance P a (0) Forgetting factor parameter lambda a0 、λ a1 、η a 。
And a second step of: and calculating a forgetting factor. Forgetting factor lambda at k moment of sampling voltage least square model a (k) The calculation formula of (2) is as follows:
λ a (k)＝λ a1 -(λ a1 -λ a0 )exp(-η a kΔt)
and a third step of: an observation matrix is calculated. Observation matrix H at k moment of sampling voltage least square model a (k) The calculation formula of (2) is as follows:
wherein, the liquid crystal display device comprises a liquid crystal display device,representing parameter a' 2 ,a' 3 An estimate at time k-1.
Fourth step: a gain matrix is calculated. Gain matrix K at sampling voltage least square model K moment a (k) The calculation formula of (2) is as follows:
fifth step: and updating an error covariance matrix and a parameter estimation value. Error covariance matrix P of sampling voltage least square model a (k) And parameter estimation valueThe calculation formula of (2) is as follows:
P a (k)＝[1-K a (k)H a (k)]P a (k-1)/λ a (k)
wherein, the liquid crystal display device comprises a liquid crystal display device,the parameter estimate at time k-1 is shown.
Sixth step: and judging whether the algorithm converges or not. Judging whether an algorithm for identifying the sampling voltage least square model converges or not, wherein the algorithm specifically comprises the following steps: in the fifth step, calculating the estimated value of the parameter at the k momentThen, the parameter a in the sliding window w is calculated 1 Mean value of estimated valuesAnd standard deviation->The calculation formula is as follows:
and then compareIf the error covariance matrix P is smaller than the convergence accuracy epsilon, if yes, the algorithm is considered to be converged, otherwise, the error covariance matrix P obtained by calculation in the fifth step is calculated a (k) And parameter estimation value->As an initial value in the first step, the second step is returned to iterate until the algorithm converges. When the algorithm converges, the mean value in the sliding window w is +.>As model parameter a 1 Is a result of the identification of (a).
In the step S5 described above: bus voltage plateau value b 1 The following ten steps are taken.
The first step: initializing. Setting initial values of parameters of a first least squares model of bus voltageInitial value of error covariance P b1 (0) Forgetting factor parameter->η b1 The method comprises the steps of carrying out a first treatment on the surface of the Parameter initial value of second least squares model for setting busbar voltage +.>Initial value of error covariance P b2 (0) Forgetting factor parameter->η b2 。
And a second step of: calculating forgetting factor lambda of first least squares model of bus voltage at k moment b1 (k) The calculation formula is as follows:
and a third step of: an observation matrix H of a first least square model for calculating the bus voltage at the moment k b1 (k) The calculation formula is as follows:
H b1 (k)＝U pn (t k-1 )-U pn (t k-2 )
fourth step: calculating gain matrix K of first least square model of bus voltage at moment K b1 (k) The calculation formula is as follows:
K b1 (k)＝P b1 (k-1)H b1 (k)/[λ b1 (k)+H b1 (k)P b1 (k-1)H b1 (k)]
fifth step: updating the error covariance P of the first least squares model of the busbar voltage at time k b1 (k) And parameter estimation valueThe calculation formula is as follows:
P b1 (k)＝[1-K b1 (k)H b1 (k)]P b1 (k-1)/λ b1 (k)
sixth step: calculating forgetting factor lambda of second least squares model of bus voltage at k moment b2 (k) The calculation formula is as follows:
seventh step: an observation matrix H of a second least square model for calculating the bus voltage at the moment k b2 (k) The calculation formula is as follows:
eighth step: bus for calculating k momentGain matrix K of second least square model of voltage b2 (k) The calculation formula is as follows:
K b2 (k)＝P b2 (k-1)H b2 (k)/[λ b2 (k)+H b2 (k)P b2 (k-1)H b2 (k)]
ninth step: updating the error covariance P of the second least squares model of the busbar voltage at time k b2 (k) And parameter estimation valueThe calculation formula is as follows:
P b2 (k)＝[1-K b2 (k)H b2 (k)]P b2 (k-1)/λ b2 (k)
tenth step: judging whether an identification algorithm of the bus voltage is converged or not, wherein the identification algorithm specifically comprises the following steps: in the ninth step, the parameter b is calculated 1 Estimated value at time kThen, calculating the parameter b in the sliding window w 1 Mean value of estimated values>And standard deviation->The calculation formula is as follows:
and then compareIf the error covariance P is smaller than the convergence accuracy epsilon, if so, the algorithm is considered to be converged, otherwise, the error covariance P obtained by the calculation in the ninth step is calculated b2 (k) And parameter estimation value->As an initial value in the first step, the second step is returned to iterate until the algorithm converges. When the algorithm converges, the mean value in the sliding window w is +.>As model parameter b 1 I.e. a plateau value of the bus voltage.
When the pulse signal generator generates a pulse signal with the amplitude of U in the positive half period or the negative half period s Respectively identifying the stable value a of the sampling voltage by the steps 1 And a plateau value b of the bus voltage 1 。
Finally, calculating the insulation resistance value of the positive electrode of the power battery relative to the electric chassis of the automobile, wherein the calculation formula is as follows:
R p ＝R(2R f (U s -a 1 )/(a 1 R+b 1 R f )-1)
the insulation resistance value of the power battery cathode relative to the automobile electric chassis is calculated, and the calculation formula is as follows:
R n ＝R(2R f (U s -a 1 )/(a 1 R-b 1 R f )-1)
wherein R is p The positive insulation resistance value of the electric automobile; r is R n The negative electrode insulation resistance value of the electric automobile; r is the resistance value of a current-limiting resistor in the insulation detection equivalent circuit; r is R f Is a sampling resistor; u (U) s Is the voltage value generated by the pulse signal generator in the positive and negative half periods.
Referring to fig. 3, a flowchart is implemented by an insulation resistance detection algorithm based on a forgetting factor recursive least square when an electric vehicle power battery is disconnected.
The method mainly comprises three parts of identification of a stable value of the sampling voltage, identification of a stable value of the bus voltage and calculation of an insulation resistance value. And firstly setting parameters w and epsilon for judging algorithm convergence when the algorithm starts, and judging whether the algorithm reaches convergence after each iteration is ended or not, and stopping iteration if the algorithm reaches convergence. The identification of the sampled voltage stationary value and the identification of the bus voltage stationary value use the same parameters converged by the judgment algorithm, namely the same sliding window w and the same convergence precision epsilon. When the algorithm runs, the identification of the sampling voltage stable value and the identification of the bus voltage stable value are performed simultaneously.
The identification of the stable value of the sampling voltage mainly comprises the following three steps:
the first step: initialization mainly comprising setting initial values of least squares parameters of sampled voltagesInitial value of error covariance P a (0) Forgetting factor parameter lambda a0 、λ a1 、η a . The least squares parameter initial value and the error covariance initial value are used for the first iteration of the recursive algorithm, and the forgetting factor parameter is used for calculating the forgetting factor of each iteration.
And a second step of: parameter identification mainly comprising forgetting factor lambda a Is calculated, observed matrix H a Is calculated, gain matrix K a Is calculated, the voltage U is sampled f An error covariance matrix and an update of parameter estimates. This step is an iterative process of the variable forgetting factor recursive least squares algorithm.
And a third step of: judging whether the algorithm converges or not, wherein the algorithm mainly comprises a parameter a in a sliding window w 1 And (3) calculating the mean value and the standard deviation of the estimated values, and comparing the standard deviation with epsilon. After each iteration is completed, the sliding window w is calculatedIf the standard deviation is less than epsilon, the algorithm is considered to converge and proceeds to the next step. Otherwise, taking the error covariance matrix and the parameter estimation value updated in the last step as the initial valuesThe values continue to iterate until the algorithm converges. When the algorithm converges, taking the average value in the sliding window as a model parameter a 1 I.e. a stationary value of the sampled voltage.
The identification of the stable value of the bus voltage mainly comprises the following four steps:
the first step: initializing. Mainly comprises setting parameter initial value of first least square model of bus voltageInitial value of error covariance P b1 (0) Forgetting factor parameter->η b1 The method comprises the steps of carrying out a first treatment on the surface of the Parameter initial value of second least squares model for setting busbar voltage +.>Initial value of error covariance P b2 (0) Forgetting factor parameter->η b2 . The least squares parameter initial value and the error covariance initial value are used for the first iteration of the recursive algorithm, and the forgetting factor parameter is used for calculating the forgetting factor of each iteration.
And a second step of: a first least squares model of the bus voltage is identified. Mainly comprises forgetting factor lambda b1 Is calculated, observed matrix H b1 Is calculated, gain matrix K b1 Is calculated, bus voltage U pn Is input to the bus voltage difference DeltaU between adjacent two times pn (k) And (3) updating the error covariance and the parameter estimation value, and inputting the updated parameter estimation value as the next step. This step is the iterative process of the first variable forgetting factor recursive least squares algorithm.
And a third step of: a second least squares model of the bus voltage is identified. Mainly comprises forgetting factor lambda b2 Is calculated, observed matrix H b2 Is calculated, gain matrix K b2 Is calculated, bus voltage U pn Is used for updating the error covariance and the parameter estimation value. This step is the iterative process of the second variable forgetting factor recursive least squares algorithm.
Fourth step: judging whether the identification of the stable value of the bus voltage is converged or not, wherein the identification mainly comprises a parameter b in a sliding window w 1 And (3) calculating the mean value and the standard deviation of the estimated values, and comparing the standard deviation with epsilon. After the third iteration is completed, the sliding window w is calculatedIf the standard deviation is less than epsilon, the algorithm is considered to converge and proceeds to the next step. Otherwise, the error covariance and the parameter estimation value updated in the second step and the third step are used as initial values to continue iteration until the algorithm converges. When the algorithm converges, taking the average value in the sliding window as a model parameter b 1 I.e. a plateau value of the bus voltage.
And when the identification of the stable value of the sampling voltage and the identification of the stable value of the bus voltage are converged, calculating the insulation resistance values of the positive electrode and the negative electrode.
The insulation resistance detection method for the disconnection of the power battery of the electric automobile provided by the embodiment of the invention can accurately and rapidly calculate the insulation resistance of the anode and the cathode. Under the condition that the power battery of the electric automobile is disconnected, the influence of the system Y capacitance on the measurement process is considered, a function model of sampling voltage and bus voltage is established, the function model is simplified, a model containing five parameters is simplified into a model with only three parameters, and the simplified bus voltage function model is subjected to parameter decoupling, so that the calculated amount is greatly reduced. And identifying the stable value of the sampling voltage and the stable value of the bus voltage by adopting a forgetting factor recursive least square algorithm, and finally calculating the insulation resistance value of the positive electrode and the negative electrode according to the identification result, thereby eliminating the influence of the Y capacitor on the insulation resistance value detection in the electric automobile and reducing the interference of noise on the sampling voltage signal and the bus voltage signal.
Furthermore, the invention simplifies the function model of the sampling voltage and the bus voltage, parameter decoupling is carried out on the simplified bus voltage function model, and recursive least square is adopted, so that the invention has the advantages of small calculated amount and small occupied memory, and is more suitable for being used in a vehicle-mounted embedded system. And the variable forgetting factor is adopted, so that the error between the result obtained by identifying the simplified model and the actual result is reduced, and the method has the advantages of high convergence speed and high convergence precision. Finally, by setting a proper basis for judging convergence, the invention can improve the response speed of the detection algorithm, avoid the waste of calculation resources and reduce errors. The invention actually considers the existence of the Y capacitor in the electric automobile, can calculate the insulation resistance value of the positive electrode and the negative electrode at the same time, and has more practicability.
The foregoing has shown and described the basic principles and main features of the present invention and the advantages of the present invention. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, and that the above embodiments and descriptions are merely illustrative of the principles of the present invention, and various changes and modifications may be made without departing from the spirit and scope of the invention, which is defined in the appended claims. The scope of the invention is defined by the appended claims and equivalents thereof.
Claims (3)
1. A method for detecting insulation resistance value when an electric automobile power battery is disconnected is characterized by comprising the following steps: comprising the following steps:
s1, constructing an insulation detection circuit model when a power battery with a Y capacitor is disconnected; the insulation detection circuit model includes: an insulation detection equivalent circuit and an electric automobile high-voltage equivalent system;
s2, generating square wave signals by an insulation detection equivalent circuit, injecting the square wave signals into an automobile high-voltage equivalent system through an electric chassis, and passing through a positive-negative insulation resistor R of an electric automobile p 、R n And positive and negative Y capacitance C p 、C n Flowing back to the insulation detection equivalent circuit;
s3, establishing a function model of the sampling voltage and the bus voltage according to the reflowed signals;
s4, simplifying the function models of the sampling voltage and the bus voltage respectively, simplifying the model with five parameters into a model with three parameters, and performing parameter decoupling on the simplified bus voltage model;
s5, identifying the sampling voltage by adopting a forgetting factor recursive least square algorithm; identifying the decoupled bus voltage model by adopting a twice forgetting factor recursive least square algorithm;
s6, calculating insulation resistance values of the positive electrode and the negative electrode according to the identification result;
wherein, the electric automobile high voltage equivalent system in S1: the power battery comprises a power battery, an anode insulation resistor, a cathode insulation resistor, an anode Y capacitor, a cathode Y capacitor and an electric chassis, wherein the anode of the power battery is electrically connected with the anode insulation resistor and the anode Y capacitor through a switch respectively, and the switch is in an off state; the negative electrode of the power battery is electrically connected with the negative electrode insulation resistor and the negative electrode Y capacitor respectively, and the positive electrode insulation resistor, the negative electrode insulation resistor, the positive electrode Y capacitor and the negative electrode Y capacitor are electrically connected with the electric chassis;
the insulation detection circuit: the power battery comprises a first current limiting resistor, a second current limiting resistor, a sampling resistor and a pulse signal generator, wherein one end of the first current limiting resistor is electrically connected with the positive electrode of the power battery through a switch, and the switch is in an off state; one end of the second current limiting resistor is electrically connected with the negative electrode of the power battery, the other end of the first current limiting resistor and the other end of the second current limiting resistor are electrically connected with the sampling resistor, the pulse signal generator is electrically connected with the electric chassis, and the pulse signal generator and the sampling resistor are grounded;
the step S4 includes:
s41, based on the sampling voltage value a when the voltage is required to reach stability when the insulation resistance value is calculated 1 And bus voltage value b 1 And other parameters not related to the calculation of insulation resistance value, the voltage U is sampled f And bus voltage U pn Simplifying the function model of (2);
s42, sampling voltage U f The functional model at time t is simplified as:
U f (t)＝a 1 +a' 2 exp(-t/a' 3 ) (3)
bus voltage U pn The functional model at time t is simplified as:
U pn (t)＝b 1 +b' 2 exp(-t/b' 3 ) (4)
wherein a is 1 And b 1 To simplify the steady-state component of the response of the post-model, i.e. a 1 B is the stationary value of the sampled voltage 1 Is a stable value of the bus voltage; a' 2 And b' 2 To simplify the response gain of the post-model, a' 3 And b' 3 Response time constants for the simplified back model; a' 2 、b' 2 、a' 3 And b' 3 The insulation resistance value is not calculated;
s43, sampling voltage U f Converting the continuous time function model of the simplified formula (3) into a discrete time function model, and performing parameter decoupling on the formula (4);
in the step S43, the sampling voltage U is sampled f The continuous time function model of the simplified formula (3) is converted into a discrete time function model, and the method is specifically expressed as follows:
U f (k)＝a 1 (k)+a' 2 (k)exp(-kΔt/a' 3 (k)) (3-1)
wherein Δt is a sampling time interval; sampling voltage U f The discrete time function model of (2) is a nonlinear model, and is linearized by adopting first-order Taylor series expansion and converted into a least square model;
in the step S43, parameter decoupling is performed on the formula (4), including:
according to the bus voltage U pn The simplified function model at time t yields U pn At t k Equation of time:
U pn (t k )＝b 1 +b' 2 exp(-t k /b' 3 ) (4-1)
then, the process is carried out,
U pn (t k )＝b 1 +U pn (t k-1 )exp(-Δt/b' 3 )-b 1 exp(-Δt/b' 3 ) (4-2)
wherein Δt is the sampling interval time; u of adjacent two moments pn The expression of (2) is subtracted to obtain the following expression:
ΔU pn (t k )＝exp(-Δt/b' 3 )*ΔU pn (t k-1 ) (4-3)
wherein DeltaU pn (t k ) At t k U at moment of time pn Subtracting t k-1 U at moment of time pn ，ΔU pn (t k-1 ) At t k-1 U at moment of time pn Subtracting t k-2 U at moment of time pn ；
Bus voltage U pn Simplified function model parameters b 1 And parameter b' 3 The decoupling is obtained, and the two least squares models after decoupling are respectively as follows: the first least squares model of the bus voltage is:
ΔU pn (k)＝H b1 (k)*β(k)+e b1 (k) (4-4)
wherein H is b1 (k)＝U pn (t k-1 )-U pn (t k-2 )，β(k)＝exp(-Δt/b 3 '(k))，ΔU pn (k)＝U pn (t k )-U pn (t k-1 )，e b1 (k) A Gaussian white noise sequence which is one-dimensional random observation noise and has zero mean value and is independent;
the second least squares model of the bus voltage is:
U pn (k)＝H b2 (k)*b 1 (k)+Y b (k)+e b2 (k) (4-5)
wherein U is pn (k) Representing t k U at moment of time pn ， E is the state estimation value at time k in the first least square model of the bus voltage b2 (k) Is thatOne-dimensional random observation noise is zero-mean value and independent Gaussian white noise sequence;
in the step S5, a variable forgetting factor recursive least square algorithm is adopted to identify the sampling voltage model, which includes:
s51, setting a least square parameter initial value, an error covariance initial value and a forgetting factor parameter of the sampling voltage; the least square parameter initial value and the error covariance initial value are used for the first iteration of the recursive algorithm, and the forgetting factor parameter is used for calculating the forgetting factor of each iteration;
s52, parameter identification, including calculation of forgetting factors, calculation of observation matrixes, calculation of gain matrixes, input of sampling voltage, error covariance matrixes and parameter estimation values a 1 Is updated according to the update of (a);
s53, calculating the parameter estimation value a in the sliding window 1 Mean and standard deviation of (a); when the standard deviation is smaller than a first preset threshold value, the parameter estimation value a 1 The average value of (2) is taken as the stable value of the sampling voltage;
in the step S5, the identification of the decoupled bus voltage model by adopting a two-time forgetting factor recursive least square algorithm includes:
identifying a first least square model of the busbar voltage to obtain an estimated value of the parameter beta; substituting the identified parameter beta into a second least square model of the bus voltage, and continuously identifying the second least square model to obtain a parameter b 1 Is a function of the estimated value of (2);
the parameter b is subjected to a two-time forgetting factor recursive least square algorithm 1 Identifying, including:
s501, respectively setting a first least square model, a parameter initial value, an error covariance initial value and a forgetting factor parameter of a bus voltage; the least square parameter initial value and the error covariance initial value are used for the first iteration of the recursive algorithm, and the forgetting factor parameter is used for calculating the forgetting factor of each iteration;
s502, identifying a first least square model of the busbar voltage to obtain a parameter estimation value beta; the parameter estimation value beta is input in the next step;
s503, identifying a second least square model of the bus voltage to obtain a parameter estimated value b 1 ；
S504, calculating the parameter estimation value b in the sliding window 1 Mean and standard deviation of (a); when the standard deviation is smaller than a second preset threshold value, the parameter estimation value b 1 The average value of (2) is taken as the stable value of the bus voltage;
in the step S6, the insulation resistance of the positive electrode and the negative electrode is calculated according to the identification result, including:
the insulation resistance value of the positive electrode of the power battery relative to the automobile electric chassis is calculated, and the calculation formula is as follows:
R p ＝R(2R f (U s -a 1 )/(a 1 R+b 1 R f )-1)
the insulation resistance value of the power battery cathode relative to the automobile electric chassis is calculated, and the calculation formula is as follows:
R n ＝R(2R f (U s -a 1 )/(a 1 R-b 1 R f )-1)
wherein R is p The positive insulation resistance value of the electric automobile; r is R n The negative electrode insulation resistance value of the electric automobile; r is the resistance value of a current-limiting resistor in the insulation detection equivalent circuit; r is R f Is a sampling resistor; u (U) s Is the voltage value generated by the pulse signal generator in the positive and negative half periods.
2. The insulation resistance detection method when the power battery of the electric automobile is disconnected according to claim 1, wherein the insulation resistance detection method comprises the following steps: the step S3 includes:
from the signal flowing back, a sampling voltage U is established f The function of the value of (c) at time t is:
establishing bus voltage U pn The function of the value of (c) at time t is:
a 1 、a 2 、a 3 、a 4 、a 5 、b 1 、b 2 、b 3 、b 4 、b 5 is an unknown parameter of the two function models, and e is a constant.
3. The insulation resistance detection method when the power battery of the electric automobile is disconnected according to claim 1, wherein the insulation resistance detection method comprises the following steps: sampling voltage U f The discrete time function model of (2) is a nonlinear model, and is linearized by adopting first-order Taylor series expansion and converted into a least square model; comprising the following steps:
u is set to f (k) Expanding the parameter estimation value around the last moment into a Taylor series, and omitting secondary and above items;
sampling voltage U f The simplified function model is converted into a least square model, and the expression is as follows:
U f (k)＝H a (k)A(k)+Y a (k)+e a (k) (3-2)
wherein H is a (k) An observation matrix at the moment k; a (k) is a parameter value at time k, and a (k) = [ a ] 1 (k),a' 2 (k),a' 3 (k)]，Y a (k) E is a constant error term a (k) Is a one-dimensional random observation noise and is a zero-mean value independent Gaussian white noise sequence.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110932123.9A CN113655281B (en) | 2021-08-13 | 2021-08-13 | Insulation resistance value detection method during disconnection of electric automobile power battery |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110932123.9A CN113655281B (en) | 2021-08-13 | 2021-08-13 | Insulation resistance value detection method during disconnection of electric automobile power battery |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113655281A CN113655281A (en) | 2021-11-16 |
| CN113655281B true CN113655281B (en) | 2023-09-26 |
Family
ID=78479857
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110932123.9A Active CN113655281B (en) | 2021-08-13 | 2021-08-13 | Insulation resistance value detection method during disconnection of electric automobile power battery |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113655281B (en) |
Citations (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2010197093A (en) * | 2009-02-23 | 2010-09-09 | Hitachi Automotive Systems Ltd | State determination device |
| JP2014153090A (en) * | 2013-02-05 | 2014-08-25 | Yazaki Corp | Insulation state detection device |
| CN105244180A (en) * | 2015-11-06 | 2016-01-13 | 海南师范大学 | Preparation of three-dimensional graphene manganese dioxide nano-composite modified electrode and capacitive property test method thereof |
| JP2016099323A (en) * | 2014-11-26 | 2016-05-30 | 矢崎総業株式会社 | Insulation state detector |
| CN108445418A (en) * | 2018-05-17 | 2018-08-24 | 福建省汽车工业集团云度新能源汽车股份有限公司 | A kind of battery dump energy evaluation method and storage medium |
| JP2019113431A (en) * | 2017-12-25 | 2019-07-11 | 矢崎総業株式会社 | Ground fault detector |
| CN110286332A (en) * | 2019-06-17 | 2019-09-27 | 南通大学 | A kind of electric automobile power battery SOC estimation method theoretical based on more new breaths |
| CN110346641A (en) * | 2018-04-08 | 2019-10-18 | 上海汽车集团股份有限公司 | A kind of insulation resistance detection system and method |
| CN110398633A (en) * | 2019-09-03 | 2019-11-01 | 北京新能源汽车技术创新中心有限公司 | A kind of insulation impedance detection method of electric car |
| CN111323705A (en) * | 2020-03-19 | 2020-06-23 | 山东大学 | Battery parameter identification method and system based on robust recursive least squares |
| CN111781424A (en) * | 2020-07-14 | 2020-10-16 | 中国第一汽车股份有限公司 | Method and device for measuring insulation resistance of electric vehicle, vehicle and storage medium |
| US10928461B1 (en) * | 2019-08-23 | 2021-02-23 | Stafl Systems, LLC | Location-determinant fault monitoring for battery management system |
| CN112816879A (en) * | 2021-01-05 | 2021-05-18 | 南京航空航天大学 | Online estimation method for power battery SoE for variable-working-condition vehicle |
| WO2021156282A1 (en) * | 2020-02-03 | 2021-08-12 | Volkswagen Aktiengesellschaft | Method for monitoring y-capacitors |
| CN113655282A (en) * | 2021-08-13 | 2021-11-16 | 海南师范大学 | Insulation resistance value detection method during connection of power battery of electric automobile |
| CN113655280A (en) * | 2021-08-13 | 2021-11-16 | 海南师范大学 | Insulation resistance value detection method during connection of power battery of electric automobile |
| CN116148536A (en) * | 2023-03-28 | 2023-05-23 | 湖南大学 | Electric automobile insulation resistance value detection method based on unbalanced bridge method |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7612532B2 (en) * | 2005-06-21 | 2009-11-03 | Gm Global Technology Operations, Inc. | Method for controlling and monitoring using a state estimator having variable forgetting factors |
| US9696384B2 (en) * | 2012-08-24 | 2017-07-04 | GM Global Technology Operations LLC | High voltage bus-to-chassis isolation resistance and Y-capacitance measurement |
-
2021
- 2021-08-13 CN CN202110932123.9A patent/CN113655281B/en active Active
Patent Citations (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2010197093A (en) * | 2009-02-23 | 2010-09-09 | Hitachi Automotive Systems Ltd | State determination device |
| JP2014153090A (en) * | 2013-02-05 | 2014-08-25 | Yazaki Corp | Insulation state detection device |
| JP2016099323A (en) * | 2014-11-26 | 2016-05-30 | 矢崎総業株式会社 | Insulation state detector |
| CN105244180A (en) * | 2015-11-06 | 2016-01-13 | 海南师范大学 | Preparation of three-dimensional graphene manganese dioxide nano-composite modified electrode and capacitive property test method thereof |
| JP2019113431A (en) * | 2017-12-25 | 2019-07-11 | 矢崎総業株式会社 | Ground fault detector |
| CN110346641A (en) * | 2018-04-08 | 2019-10-18 | 上海汽车集团股份有限公司 | A kind of insulation resistance detection system and method |
| CN108445418A (en) * | 2018-05-17 | 2018-08-24 | 福建省汽车工业集团云度新能源汽车股份有限公司 | A kind of battery dump energy evaluation method and storage medium |
| CN110286332A (en) * | 2019-06-17 | 2019-09-27 | 南通大学 | A kind of electric automobile power battery SOC estimation method theoretical based on more new breaths |
| US10928461B1 (en) * | 2019-08-23 | 2021-02-23 | Stafl Systems, LLC | Location-determinant fault monitoring for battery management system |
| CN110398633A (en) * | 2019-09-03 | 2019-11-01 | 北京新能源汽车技术创新中心有限公司 | A kind of insulation impedance detection method of electric car |
| WO2021156282A1 (en) * | 2020-02-03 | 2021-08-12 | Volkswagen Aktiengesellschaft | Method for monitoring y-capacitors |
| CN111323705A (en) * | 2020-03-19 | 2020-06-23 | 山东大学 | Battery parameter identification method and system based on robust recursive least squares |
| CN111781424A (en) * | 2020-07-14 | 2020-10-16 | 中国第一汽车股份有限公司 | Method and device for measuring insulation resistance of electric vehicle, vehicle and storage medium |
| CN112816879A (en) * | 2021-01-05 | 2021-05-18 | 南京航空航天大学 | Online estimation method for power battery SoE for variable-working-condition vehicle |
| CN113655282A (en) * | 2021-08-13 | 2021-11-16 | 海南师范大学 | Insulation resistance value detection method during connection of power battery of electric automobile |
| CN113655280A (en) * | 2021-08-13 | 2021-11-16 | 海南师范大学 | Insulation resistance value detection method during connection of power battery of electric automobile |
| CN116148536A (en) * | 2023-03-28 | 2023-05-23 | 湖南大学 | Electric automobile insulation resistance value detection method based on unbalanced bridge method |
Non-Patent Citations (4)
| Title |
|---|
| A new method of insulation detection on electric vehicles based on a variable forgetting factor recursive least squares algorithm;Zhenbin Chen 等;《Journal & Magazines》;第9卷;73590-73607 * |
| 基于模型参数辨识的电动汽车绝缘电阻测量;张强;石春;;电子测量与仪器学报(11);83-86 * |
| 电动汽车动力电池绝缘性能监测方法;刘少尧;《中国优秀硕士学位论文全文数据库工程科技II辑》(第1期);C035-1778 * |
| 电阻绝缘检测中低压脉冲信号注入法的算法改进;黄雨龙 等;《海南大学学报自然科学版》;第38卷(第1期);13-19 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113655281A (en) | 2021-11-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113655277B (en) | Insulation resistance value detection method during disconnection of electric automobile power battery | |
| WO2021259196A1 (en) | Battery pack consistency evaluation method and system | |
| CN106842060A (en) | A kind of electrokinetic cell SOC estimation method and system based on dynamic parameter | |
| CN104181470B (en) | Battery state-of-charge (SOC) estimation method based on nonlinear prediction extended Kalman filtering | |
| CN103616647B (en) | A kind of estimation method of battery dump energy for cell management system of electric automobile | |
| Hung et al. | State-of-charge and state-of-health estimation for lithium-ion batteries based on dynamic impedance technique | |
| CN113655280B (en) | Insulation resistance value detection method during connection of power batteries of electric vehicles | |
| CN111581904B (en) | Lithium battery SOC and SOH collaborative estimation method considering cycle number influence | |
| CN109870651A (en) | A kind of electric automobile power battery system SOC and SOH joint estimation on line method | |
| CN107741569A (en) | A kind of evaluation method of the lithium battery charge state based on segment extension Kalman filtering | |
| CN109581225A (en) | The energy state evaluation method and battery management system of battery on-line parameter identification | |
| CN108761340A (en) | The battery evaluation method of strong tracking volume Kalman filtering based on noise jamming | |
| Huang et al. | Soc estimation of li-ion battery based on improved ekf algorithm | |
| CN113311239B (en) | Electric automobile insulation resistance detection circuit and method | |
| CN106526488A (en) | Fault diagnosis method of sensors in tandem type power battery pack | |
| CN108828448B (en) | Battery state of charge on-line estimation method based on charging voltage curve fusion Kalman filtering | |
| CN116148536B (en) | Electric automobile insulation resistance value detection method based on unbalanced bridge method | |
| CN113655282A (en) | Insulation resistance value detection method during connection of power battery of electric automobile | |
| CN112305440A (en) | Battery remaining capacity and health state joint estimation method | |
| CN110673037B (en) | Battery SOC estimation method and system based on improved simulated annealing algorithm | |
| CN113156321A (en) | Estimation method for state of charge (SOC) of lithium ion battery | |
| CN113655279B (en) | Insulation resistance value detection method during disconnection of electric automobile power battery | |
| CN113655278B (en) | Insulation resistance value detection method during connection of power batteries of electric vehicles | |
| CN113655281B (en) | Insulation resistance value detection method during disconnection of electric automobile power battery | |
| Baba et al. | State of charge estimation of lithium-ion battery using Kalman filters |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |