Metering method for concrete mixing ship to eliminate wave excitation error

Abstract

The present invention provides a concrete mixing ship metering method for eliminating wave excitation errors, the method comprising: S1, determining the STFT window size; S2, establishing a system model; S3, Kalman filtering; S4, STFT solving the signal period; S5, sliding average filtering; S6, repeating steps 3 to 5. By processing the real signal, random error and wave excitation error contained in the output signal of the weighing sensor, the periodic error and random error caused by external wave excitation in the output signal of the weighing sensor are successfully identified and eliminated, and the metering accuracy of the concrete mixing ship metering system is fully improved.

Description

Technical Field

[0001] The invention relates to the field of offshore material measurement, and in particular to a concrete mixing ship measurement method for eliminating wave excitation errors. Background Art

[0002] With the development of marine infrastructure construction, ship-mounted concrete mixing plants have been widely used. For ship-mounted concrete mixing plants, ingredient metering is the core function with the highest precision requirements. When a ship-mounted concrete mixing plant is operating under wave excitation, the ship will produce a compound motion of heave, roll, pitch, sway and surge due to the excitation of the waves, which will cause its metering system to be inaccurate.

[0003] The operating principle of the weighing sensor is to detect the pressure or tension of the weighing object on the weighing sensor. When the weighing is static, the weighing object and the weighing sensor are in a relatively static state. At this time, the weighing result is relatively accurate, and its error is only related to the accuracy of the weighing sensor. When the ship-mounted concrete mixing plant is under the action of wave excitation, the weighing sensor will detect the heave acceleration, sway acceleration, pitch acceleration, roll angular acceleration and pitch angular acceleration of the weighing object caused by the wave excitation, that is, there will be relative movement between the weighing object and the weighing sensor, which will cause the weight of the weighing object detected by the weighing sensor to change in real time. In the above process, the weighing accuracy of the weighing sensor cannot meet the requirements of the relevant specifications for concrete production.

[0004] At present, Chinese patent CN112985557A proposes a material metering device for ships, which can realize high-precision metering of materials on ships. Solid materials (aggregates, powders) or liquid materials can eliminate the influence of surge on the weighing sensor through the correction algorithm of the correction device, thereby effectively improving the accuracy of the material metering system of concrete mixing ships. However, this patent realizes accurate measurement by calculating between the weight measurement value of the material and the weight measurement value of the standard mass block, which requires the deployment of multiple sensors and corresponding load-bearing equipment, and the structure is complex. At the same time, it is easy to receive external interference during implementation, resulting in inaccurate measurement. Therefore, a concrete mixing ship metering method that eliminates wave excitation errors is proposed to solve the above problems. Summary of the invention

[0005] The main purpose of the present invention is to provide a concrete mixing ship metering method for eliminating wave excitation errors, which can effectively eliminate the periodic errors caused by periodic wave excitation, thereby improving the weighing accuracy and ensuring the concrete batching metering accuracy.

[0006] In order to solve the above technical problems, the technical solution adopted by the present invention is: a concrete mixing ship metering method for eliminating wave excitation errors, the method comprising:

[0007] S1, determine the STFT window size, and determine the size of the STFT sliding window based on prior data;

[0008] S2, establish a system model, and construct a weighing system model according to step 1;

[0009] S3, Kalman filtering, performs traditional Kalman filtering through the constructed weighing system model;

[0010] S4, STFT solves the signal period, performs short-time Fourier transform on the signal that passes through the traditional Kalman filter, and determines the signal period caused by wave excitation by analyzing its energy spectrum distribution diagram;

[0011] S5, sliding average filtering, determining the size of the sliding window by the period obtained in step 4 and the signal acquisition frequency, and then performing sliding window mean filtering on the signal data within the period;

[0012] S6. Repeat steps 3 to 5.

[0013] In a preferred embodiment, the length of data included in the STFT window is determined based on prior data of the wave period in the sea area where the ship is located.

[0014] In a preferred embodiment, the weighing system model includes observation equations and state equations of the weighing system under discrete state interference caused by random errors and wave excitations, and initializes the initial state vector and error covariance matrix.

[0015] In the preferred embodiment, the state equation is:

[0016] X(t)=AX(t-1)+R

[0017] In the formula, X(t) and X(t-1) are the state variables at time t and time t-1 respectively, R is the process noise vector, R is Gaussian white noise, and satisfies R~(0,Q), where Q is the covariance matrix of the process noise vector R.

[0018] In the preferred embodiment, the observation equation is:

[0019] Z(t)=CX(t)+W

[0020] Where Z(t) is the observation at time t, W is the observation noise vector, W is Gaussian white noise, and satisfies W~(0,E), and E is the covariance matrix of the observation noise vector W.

[0021] In the preferred embodiment, the traditional Kalman filtering process is divided into two stages: a prediction stage and an update stage.

[0022] In the preferred solution, the implementation process of the prediction stage is:

[0023] Status prediction:

[0024] X(t|t-1)=AX(t-1)

[0025] Error covariance prediction:

[0026] P(t|t-1)=AP(t-1)A T +Q

[0027] The implementation process of the update phase is:

[0028] Kalman gain update:

[0029]

[0030] Error covariance update:

[0031] P(t)=(IK(t)C)P(t|t-1)

[0032] Status Update:

[0033] X(t)=X(t|t-1)+K(t)(Z(t)-CX(t|t-1))

[0034] Where X(t) is the state value at time t, Z(t) is the observation value at time t, X(t|t-1) is the estimated state value at time t calculated based on time t-1, P(t) is the error covariance matrix at time t, P(t|t-1) is the estimated error covariance matrix at time t calculated based on time t-1, and K(t) is the Kalman gain at time t.

[0035] In the preferred embodiment, the short-time Fourier transform is:

[0036]

[0037] Where X(T) is the state value after traditional Kalman filtering at time t, and H(Tt) is the sliding window function.

[0038] In the preferred embodiment, the energy spectrum distribution of the short-time Fourier transform is:

[0039] SPEC(t,ω)=|STFT(t,ω)| 2

[0040] By analyzing the energy spectrum distribution diagram of the short-time Fourier transform, the signal frequency ω caused by wave excitation can be obtained. MAX , which is expressed as:

[0041] ω MAX =MAX(SPEC(t,ω))

[0042] In the formula, ω MAX is the value ω when SPEC(t,ω) reaches its maximum value.

[0043] In the preferred solution, the sliding window mean filtering is specifically: according to the signal acquisition frequency F and the signal frequency ω MAX Perform sliding window mean filtering, and the signal value X(t) at time t is:

[0044]

[0045] The present invention provides a concrete mixing ship metering method for eliminating wave excitation errors. By processing the real signal, random error and wave excitation error contained in the output signal of the weighing sensor, the periodic error and random error caused by external wave excitation in the output signal of the weighing sensor are successfully identified and eliminated, and the metering accuracy of the concrete mixing ship metering system is fully improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] The present invention will be further described below in conjunction with the accompanying drawings and embodiments:

[0047] Figure 1 It is a flow chart of the present invention; DETAILED DESCRIPTION

[0048] Example 1

[0049] like Figure 1 As shown, a concrete mixing ship metering method for eliminating wave excitation errors comprises:

[0050] S1. Determine the STFT window size. The size of the STFT sliding window is determined according to the prior data. The length of the data contained in the window is determined according to the prior data of the wave period of the sea area where the ship is located.

[0051] S2. Establish a system model. According to step 1, construct a weighing system model, list the observation equations and state equations of the weighing system under discrete state interference by random errors and wave excitation interference, and initialize the initial state vector and error covariance matrix;

[0052] S3, Kalman filtering, performs traditional Kalman filtering through the constructed weighing system model;

[0053] S4, STFT solves the signal period, performs short-time Fourier transform on the signal that passes through the traditional Kalman filter, and determines the signal period caused by wave excitation by analyzing its energy spectrum distribution diagram;

[0054] S5, sliding average filtering, determining the size of the sliding window by the period obtained in step 4 and the signal acquisition frequency, and then performing sliding window mean filtering on the signal data within the period;

[0055] S6. Repeat steps 3 to 5 until the measurement is completed.

[0056] Example 2

[0057] Further illustrate with reference to Example 1, Figure 1 As shown, first, a sliding window function H(t) is set, and the data length n that the window can contain is determined according to the prior data of the wave period of the sea area where the ship is located.

[0058] Secondly, assume that the state equation of the system is:

[0059] X(t)=AX(t-1)+R (1)

[0060] Z(t)=CX(t)+W (2)

[0061] In the formula, X(t) and X(t-1) are the state variables at time t and time t-1 respectively, Z(t) is the observation variable at time t, R is the process noise vector, W is the observation noise vector, R and W are both Gaussian white noise, and satisfy R~(0,Q), W~(0,E), where Q is the covariance matrix of the process noise vector R, and E is the covariance matrix of the observation noise vector W.

[0062] The traditional Kalman filter process can be divided into two stages: prediction step and update step. The implementation process of the prediction stage is:

[0063] Status prediction:

[0064] X(t|t-1)=AX(t-1) (3)

[0065] Error covariance prediction:

[0066] P(t|t-1)=AP(t-1)A T +Q (4)

[0067] The implementation process of the update phase is:

[0068] Kalman gain update:

[0069]

[0070] Error covariance update:

[0071] P(t)=(IK(t)C)P(t|t-1) (6)

[0072] Status Update:

[0073] X(t)=X(t|t-1)+K(t)(Z(t)-CX(t|t-1)) (7)

[0074] In equations (3) to (7), X(t) is the state value at time t, Z(t) is the observation value at time t, X(t|t-1) is the estimated state value at time t calculated based on time t-1, P(t) is the error covariance matrix at time t, P(t|t-1) is the estimated error covariance matrix at time t calculated based on time t-1, and K(t) is the Kalman gain at time t.

[0075] Then, perform a short-time Fourier transform on X(t):

[0076]

[0077] Where X(T) is the state value after traditional Kalman filtering at time t, and H(Tt) is the sliding window function.

[0078] The energy spectrum distribution of short-time Fourier transform is:

[0079] SPEC(t,ω)=|STFT(t,ω)| 2 (9)

[0080] By analyzing the energy spectrum distribution diagram of the short-time Fourier transform, the signal frequency ω caused by wave excitation can be obtained. MAX , which is expressed as:

[0081] ω MAX =MAX(SPEC(t,ω)) (10)

[0082] In the formula, ω MAX is the value ω when SPEC(t,ω) reaches its maximum value.

[0083] Finally, according to the signal acquisition frequency F and signal frequency ω MAX Perform sliding window mean filtering, and the signal value X(t) at time t is:

[0084]

[0085] The above embodiments are only preferred technical solutions of the present invention and should not be regarded as limiting the present invention. The protection scope of the present invention shall be the technical solutions recorded in the claims, including equivalent replacement solutions of the technical features in the technical solutions recorded in the claims. That is, equivalent replacement improvements within this scope are also within the protection scope of the present invention.

Claims

1. A concrete mixing ship metering method for eliminating wave excitation errors, characterized by: The method includes: S1, determine the STFT window size, and determine the size of the STFT sliding window based on prior data; S2, establish a system model, and construct a weighing system model according to step 1; S3, Kalman filtering, performs traditional Kalman filtering through the constructed weighing system model; S4, STFT solves the signal period, performs short-time Fourier transform on the signal that passes through the traditional Kalman filter, and determines the signal period caused by wave excitation by analyzing its energy spectrum distribution diagram; S5, sliding average filtering, determining the size of the sliding window by the period obtained in step 4 and the signal acquisition frequency, and then performing sliding window mean filtering on the signal data within the period; S6. Repeat steps 3 to 5.

2. The concrete mixing ship metering method for eliminating wave excitation error according to claim 1 is characterized by: The data length included in the STFT window is determined according to the priori data of the wave period in the sea area where the ship is located.

3. The concrete mixing ship metering method for eliminating wave excitation error according to claim 1 is characterized by: The weighing system model includes a weighing system observation equation and a state equation under discrete state which are disturbed by random errors and wave excitations, and initializes an initial state vector and an error covariance matrix.

4. The concrete mixing ship metering method for eliminating wave excitation error according to claim 3 is characterized by: The state equation is: X(t)=AX(t-1)+R In the formula, X(t) and X(t-1) are the state variables at time t and time t-1 respectively, R is the process noise vector, R is Gaussian white noise, and satisfies R~(0,Q), where Q is the covariance matrix of the process noise vector R.

5. The concrete mixing ship metering method for eliminating wave excitation error according to claim 4 is characterized by: The observation equation is: Z(t)=CX(t)+W Where Z(t) is the observation at time t, W is the observation noise vector, W is Gaussian white noise, and satisfies W~(0,E), and E is the covariance matrix of the observation noise vector W.

6. The concrete mixing ship metering method for eliminating wave excitation error according to claim 5 is characterized by: The traditional Kalman filtering process is divided into two stages: a prediction stage and an update stage.

7. The concrete mixing ship metering method for eliminating wave excitation error according to claim 6 is characterized by: The implementation process of the prediction stage is as follows: Status prediction: X(t|t-1)=AX(t-1) Error covariance prediction: P(t|t-1)=AP(t-1)A T +Q The implementation process of the update phase is: Kalman gain update: Error covariance update: P(t)=(IK(t)C)P(t|t-1) Status Update: X(t)=X(t|t-1)+K(t)(Z(t)-CX(t|t-1)) Where X(t) is the state value at time t, Z(t) is the observation value at time t, X(t|t-1) is the estimated state value at time t calculated based on time t-1, P(t) is the error covariance matrix at time t, P(t|t-1) is the estimated error covariance matrix at time t calculated based on time t-1, and K(t) is the Kalman gain at time t.

8. The concrete mixing ship metering method for eliminating wave excitation error according to claim 1 or 7 is characterized in that: The short-time Fourier transform is: Where X(T) is the state value after traditional Kalman filtering at time t, and H(Tt) is the sliding window function.

9. The concrete mixing ship metering method for eliminating wave excitation error according to claim 8 is characterized by: The energy spectrum distribution of the short-time Fourier transform is: SPEC(t,ω)=|STFT(t,ω)| 2 By analyzing the energy spectrum distribution diagram of the short-time Fourier transform, the signal frequency ω caused by wave excitation can be obtained. MAX , which is expressed as: ω MAX =MAX(SPEC(t,ω)) In the formula, ω MAX is the value ω when SPEC(t,ω) reaches its maximum value.

10. The concrete mixing ship metering method for eliminating wave excitation error according to claim 9 is characterized by: The sliding window mean filtering is specifically as follows: according to the signal acquisition frequency F and the signal frequency ω MAX Perform sliding window mean filtering, and the signal value X(t) at time t is:

Images