Parameter estimation method for crank balance pumping unit for electrical parameter conversion indicator diagram

Abstract

The present application relates to pumping unit device field, especially crank balance pumping unit parameter estimation method for electric parameter conversion indicator diagram, the present application is by using the corresponding motor torque of upper and lower dead point, according to parameter automatic correction algorithm, automatically corrects crank maximum balance torque, balance phase angle and motor no-load torque, solves the problem that crank maximum balance torque, balance phase angle, motor no-load torque and other parameters are difficult to obtain or inaccurate.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of pumping unit device, in particular to a parameter estimation method for crank balance pumping unit used for electric parameter conversion indicator diagram. BACKGROUND

[0002] An indicator diagram is a closed curve with the suspension point displacement as the horizontal coordinate and the suspension point load as the vertical coordinate in a stroke cycle, and contains a large amount of useful information of an oil well. Through the indicator diagram, the working conditions of the oil well can be determined, such as pump sticking, insufficient liquid supply, sucker rod breakage, valve leakage, gas influence, barrel separation, and waxing, and parameters such as stroke loss, pumping unit load utilization rate, oil well liquid production, water cut, and dynamic liquid level can be analyzed.

[0003] Due to the special geographical environment of an oilfield, the distribution range is extensive, and currently, the indicator diagram test mainly adopts the following methods:

[0004] (1) The first method is to use a portable indicator diagram instrument. This method needs to test one well after another on site by manual work periodically, and has problems of long cycle, high labor intensity, and poor timeliness, and due to the service life limitation of the nylon wire and the potentiometer, it is not suitable for online testing.

[0005] (2) The second method is to use a load sensor and an angular displacement sensor to automatically collect the suspension point load and the suspension point relative displacement, and to calculate and generate a polished rod indicator diagram to realize remote monitoring of the polished rod indicator diagram of the oil well, and to replace manual testing, but the load sensor is limited, and has problems of easy damage during well repair and being not conducive to standardization construction.

[0006] (3) The third method is to use a load displacement integrated sensor, and the displacement test uses an acceleration sensor, and for low stroke frequency oil wells, there is a problem of inaccurate testing.

[0007] (4) The fourth method is to use a three-phase electric parameter acquisition module to obtain the input power of the motor, to obtain the real-time speed of the motor through a speed sensor, and to obtain the triggering time of the upper dead point or the lower dead point through an upper dead point or lower dead point trigger, and then to derive the indicator diagram at the suspension point through a physical model, but the construction of the physical model needs the data of the pumping unit, the sucker rod, and the oil pump, and the data often has problems of being difficult to obtain, being inaccurate, or frequently changing, which limits the practical application of this method.

[0008] The fourth method needs to increase the cost and maintenance cost of the equipment the least, but the physical modeling needs many data, and has problems of being inaccurate and frequently changing. Therefore, a method needs to be found which uses less data and can automatically correct inaccurate values to ensure the accuracy of the physical modeling. SUMMARY

[0009] The application aims to provide a crank balance pumping unit parameter estimation method for electric parameter conversion indicator diagram, which can automatically correct inaccurate values.

[0010] The technical solution of the application is as follows:

[0011] The crank balance pumping unit parameter estimation method for electric parameter conversion indicator diagram is characterized in that step 1: a sampling module completes data sampling of motor torques corresponding to different positions of a horse head in a period according to a period, and sends the sampling data to an upper computer; the upper computer extracts at least 20 sample periods, and obtains motor torques corresponding to top dead centers and bottom dead centers in each period .

[0012] Step 2: according to formula 1:

[0013] estimate M cmax and τ, wherein η m is a motor efficiency which can be directly obtained from a motor manufacturer, r is a transmission ratio, M m is a motor torque, M m0 is a motor idle torque; since the output torque is much larger than the motor idle torque, M m0 = 0 in this calculation; according to formula 4: r = N / N c , calculate the transmission ratio, wherein N is a motor speed, N c is a crank speed, and the motor speed is obtained from a sensor; according to formula 5: N c = 60 / T, calculate the crank speed, wherein T is a crank period, that is, the time elapsed between two times of triggering of a crank fixed point trigger; thus, M and τ * can be obtained.

[0014] After obtaining the fixed balance phase angle τ * according to step 2, since the balance phase angle τ * is unchanged, the formula 1 is modified again to obtain formula 2:

[0015] ;

[0016] Substitute τ * obtained in step 2 into formula 2 to calculate the optimized M

[0017] ​The beneficial effect of the present application is that the present application uses a parameter automatic correction algorithm to automatically correct the maximum balance torque of the crank, the balance phase angle and the motor no-load torque, solves the problem of inaccurate parameters such as the maximum balance torque of the crank, the balance phase angle and the motor no-load torque, and the problem that the maximum balance torque of the crank often changes, makes the electric parameter conversion indicator diagram method more easy to use and operable, realizes low-cost remote monitoring of the oil well polished rod indicator diagram, and replaces manual testing. BRIEF DESCRIPTION OF DRAWINGS

[0018] The foregoing and other objects, features and advantages of the present application will become apparent to those skilled in the art from the following detailed description in conjunction with the accompanying drawings.

[0019] Wherein: Figure 1 is a schematic diagram of a structure of an oil pumping unit model of the present application;

[0020] Figure 2 is a flowchart of the present application; DETAILED DESCRIPTION

[0021] For the crank balance oil pumping unit, there is formula 1: Wherein M is the torque of the crank shaft, W is the polished rod load, B is the imbalance value of the oil pumping unit itself, is the torque factor, M cmax is the maximum balance torque of the crank, τ is the balance phase angle of the crank;

[0022] Since is a function of the crank angle θ, there is formula 2:

[0023]

[0024] Wherein Δθ is the change of the crank angle, ΔS is the change of the polished rod position, Δt is the change of time, v p is the linear velocity of the polished rod, ω c is the rotational angular velocity of the crank; since at the top and bottom dead points, the linear velocity of the polished rod is 0, therefore at the top dead point or the bottom dead point position is equal to zero, formula 3 is obtained:

[0025]

[0026] Substitute the top and bottom dead point positions of into formula 1 to obtain formula 4: M t +M cmax sin(θ t +τ)=0 and formula 5: M b +M cmax sin(θ b +τ)=0

[0027] where θ t represents the corresponding crank angle at the top dead center; θ b represents the corresponding crank angle at the bottom dead center. When the pumping unit model and stroke are determined, θ t and θ b are fixed values, which can be solved by the formula in the appendix.

[0028] In addition, the crankshaft torque is obtained from the motor torque through the transmission mechanism, and formula 6 is obtained: M = η m · r · (M m - M m0 ); where η m is the motor efficiency which can be directly obtained from the motor nameplate, r is the transmission ratio, M m is the motor torque, and M m0 is the motor idle torque. Therefore, formula 7 is obtained according to formula 4, formula 5, and formula 6: and formula 8: where, represents the motor torque corresponding to the top dead center, represents the motor torque corresponding to the bottom dead center.

[0029] The sampling module (the acquisition module can adopt a three-phase electric parameter acquisition module or an electric parameter acquisition module, and the torque formula Mm = 9550P / N, where P is the motor active power and N is the motor speed) completes the data sampling of the motor torque corresponding to different positions of the horse head in a period according to the period, and sends the sampling data to the upper computer; the upper computer extracts at least 20 sample periods, and obtains the motor torque corresponding to the top dead center and the motor torque corresponding to the bottom dead center in each period. When the motor torque corresponding to the top dead center and the bottom dead center in each sampling period is collected, the position of the bottom dead center can be detected by the proximity switch placed at the corresponding crank of the bottom dead center, so as to determine the motor torque corresponding to the bottom dead center in the period; under the premise that the crank rotates at a uniform speed, the position corresponding to the top dead center can be calculated according to the angle relationship between the top dead center and the bottom dead center, so as to determine the motor torque corresponding to the top dead center in the period. The specific mode is as follows: first, the proportional relationship of the top dead center in a period starting from the bottom dead center is calculated according to the formula α = (θ t - θ b ) / (2π), where a represents the proportion of the angle through which the crank runs from the bottom dead center to the top dead center to the angle of a period; then if q motor torques are sampled in a period, the torque of the round(a·q)th is selected as the top dead center torque. Wherein the round function represents the rounding function; the solving formula of θ b ​

[0030]

[0031] θ t The solution formula:

[0032]

[0033] Where R is the radius of the pumping unit crank, P is the length of the connecting rod, A is the length of the beam front arm, C is the length of the beam rear arm, I is the horizontal distance between the beam support center and the crank rotation center, h is the vertical distance between the beam support center and the crank rotation center, K is the distance between the beam support center and the crank rotation center, and S is the distance between the beam support center and the connecting rod shaft.

[0034] Although the upper and lower dead points satisfy formula 6 and formula 7, it is impossible to be exactly equal to 0 in actual testing, but only close to 0; in order to estimate more accurate M cmax and τ, formula 6 and formula 7 are modified to formula 8: Where f1(M cmax , τ) is the cost function, wherein:

[0035]

[0036]

[0037] Where η m is the motor efficiency, which can be directly obtained from the motor manufacturer, r is the transmission ratio, M m is the motor torque, M m0 is the motor idle torque; the transmission ratio is calculated according to formula 10: r = N c / N c , wherein N is the motor speed, and N c is the speed of the crank; the speed of the motor is obtained from the sensor; the speed of the crank is calculated according to formula 11: N c = 60 / T, wherein T is the period of the crank, that is, the time elapsed between two times of triggering of the crank fixed point trigger.

[0038] In this step, a large amount of instance data is used for optimization calculation. Since the output torque is much larger than the motor idle torque, the motor idle torque is ignored in this step, that is, N m0 = 0; the following can be obtained:

[0039]

[0040] Use the initial point: Where and τ e are arbitrary estimates.

[0041] f1(M cmas , T) is the gradient direction function of the cost function, and

[0042]

[0043] The cost function, the gradient direction function of the cost function, and the initial point are input into the BFGS algorithm, and the optimal solution of M cmax and τ is obtained; let them be M and τ * . Since the phase angle will not change after the pumping unit model is determined, τ * is considered as the estimated optimal phase angle; the maximum balance torque of the crank can be adjusted by changing the weight, so the following steps will only take as the initial value, and the initial value will still be used to further estimate the current maximum balance torque of the crank.

[0044] The balance phase angle τ * and the maximum balance torque of the crank M

[0045] The current step uses the results of the previous step and combines the current motor torque to make more refined estimates of the maximum balance torque of the crank and the motor no-load torque. The balance phase angle τ * will not change after it is determined, but the maximum balance torque of the crank can change by adjusting the number and position of the weight blocks. Therefore, before a single indicator diagram calculation, formula 8 needs to be modified to re-correct M m0 and M cmax . That is, formula 12:

[0046] Where f2(M cmax , M m0 ) is the cost function, and

[0047]

[0048]

[0049] The initial point and the gradient direction function are used:

[0050]

[0051] The cost function, the gradient direction function of the cost function, and the initial point are input into the BFGS algorithm, and the optimal solution of M and is obtained. Thus, the optimal estimated values of M and τ are obtained.* The values are brought into the formula for calculating the load of the suspension point of the beam-pumping unit, and the load of the suspension point of the horse head can be accurately obtained.

[0052] Since the formula for calculating the displacement of the suspension point and the formula for calculating the load of the suspension point are known formulas, specific description is not made herein; the formula for calculating the displacement of the suspension point is as follows:

[0053]

[0054] wherein The formula for calculating the load of the suspension point is as follows: wherein R is the radius of the crank, P is the length of the connecting rod, A is the length of the front arm of the beam, C is the length of the rear arm of the beam, I is the horizontal distance between the center of the beam support and the rotation center of the crank, h is the vertical distance between the center of the beam support and the rotation center of the crank, K is the distance between the center of the beam support and the rotation center of the crank, and S is the distance between the center of the beam support and the shaft of the connecting rod; since the formula for calculating the displacement of the suspension point and the formula for calculating the load of the suspension point are known formulas through the position of the lower dead point, the real-time rotating speed of the motor, the real-time rotating number of the motor and other data, for example, the specification of patent application No. 202110037183, patent application name: a testing device for indirectly obtaining a indicator diagram and a method thereof, paragraph 0034 to 0128; therefore, specific description is not made herein.

[0055] The above is only a preferred embodiment of the present application, and does not limit the present application in any form. Any simple modification or equivalent change made on the basis of the technical essence of the present application to the above embodiment falls within the protection scope of the present application.

Claims

1. A crank balanced pumping unit parameter estimation method for electrical parameter to performance chart, characterized in that, Step 1: the sampling module completes the data sampling of the motor torque corresponding to different positions of the horse head in a period according to the period, and sends the sampling data to the upper computer; the upper computer extracts at least 20 sample periods, and obtains the motor torque corresponding to the top dead center and the bottom dead center in each period ​ Step 2: According to Formula 1: Estimate M cmax and τ, where η m is the motor efficiency obtained directly from the motor manufacturer, r is the gear ratio, M m is the motor torque, M m0 is the motor no-load torque, θ t represents the corresponding crank angle at top dead center, M cmax is the maximum equilibrium torque of the crank, τ is the equilibrium phase angle of the crank; θ b represents the corresponding crank angle at bottom dead center; the gear ratio is calculated according to the formula: r = N / N c , where N is the motor speed, N c is the speed of the crank, the motor speed being obtained from a sensor; the speed of the crank is calculated according to the formula: N c = 60 / T, where T is the crank period, i.e. the time elapsed between two triggers of the crank fixed point trigger. Step 3: Since the output torque is much larger than the motor no-load torque, M m0 = 0 in this calculation, we get: where h is the vertical distance between the center of the beam support and the center of the crank rotation, using the initial point: where and τ e is an arbitrary estimate; The gradient direction function of the cost function f1(M cmax , τ) is: The cost function f1(M comax , τ), the gradient direction function of the cost function and the initial point are input into the BFGS algorithm, and the optimal solution of M cmax and τ is obtained; let them be M and τ * respectively; and M and τ * are obtained.

2. The crank balanced pumping unit parameter estimation method for electrical parameter-to-slug plot according to claim 1, wherein, The fixed equilibrium phase angle τ is obtained according to step 3 * After that, since the equilibrium phase angle τ * is constant, for optimization The formula of step 3 Substituting equation 1 into equation 2 gives equation 3: where f2(M cmax ,M m0 ) is a cost function, where: Using initial points and gradient direction functions: The cost function, the gradient direction function of the cost function and an initial point are input to the BFGS algorithm, and an optimal solution is obtained and

Images