Coriolis flowmeter, diagnosis method, diagnosis program and diagnosis system

Abstract

To effectively diagnose generation of an abnormal state in a vibration tube of a Coriolis flowmeter.SOLUTION: A Coriolis flowmeter 10 acquires measurement parameters related to measurement of fluid, calculates vibration tube parameters indicating a state of a vibration tube, and inputs at least one of the measurement parameters and the vibration tube parameters into a trained machine learning model that has been trained to output a diagnosis result for the vibration tube in response to the input of the parameters, thereby diagnosing generation of an abnormal state in the vibration tube.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a Coriolis flowmeter, a diagnostic method, a diagnostic program, and a diagnostic system. [Background technology]

[0002] A Coriolis flowmeter is one type of flowmeter used in plants, and is capable of measuring the flow rate and density of a fluid based on the Coriolis force of a U-shaped vibrating tube through which the fluid flows. Since a Coriolis flowmeter measures the flow rate and density by utilizing the vibration of the vibrating tube, errors in the flow rate measurement will occur if corrosion or other damage occurs in the vibrating tube. Furthermore, errors in the flow rate measurement will occur due to deposits on the inside of the vibrating tube. Conventionally, there are known techniques for diagnosing the occurrence of corrosion of the vibrating tube of a Coriolis flowmeter and the occurrence of deposits on the vibrating tube. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent No. 5340122 [Patent Document 2] Patent No. 4621140 [Patent Document 3] Patent No. 4952820 [Patent Document 4] Patent No. 7024466 Summary of the Invention [Problem to be solved by the invention]

[0004] However, with the above techniques, it is difficult to effectively diagnose the occurrence of an abnormal state in the vibrating tube of the Coriolis flowmeter, For example, it is difficult to quantitatively grasp the corrosion or adhesion of the vibrating tube.

[0005] The present invention has been made in view of the above, and an object of the present invention is to effectively diagnose the occurrence of an abnormal state in a vibrating tube of a Coriolis flowmeter. [Means for solving the problem]

[0006] The present invention provides a Coriolis flowmeter that measures the flow rate and density of a fluid flowing through a vibrating tube based on the vibration of the vibrating tube, and includes an acquisition unit that acquires measurement parameters related to the measurement of the fluid by the Coriolis flowmeter, a calculation unit that calculates vibrating tube parameters that indicate a state of the vibrating tube, and a diagnosis unit that inputs at least one of the measurement parameters and the vibrating tube parameters into a trained machine learning model that has been trained to output a diagnosis result for the vibrating tube in response to the input of parameters, and diagnoses the occurrence of an abnormal condition in the vibrating tube.

[0007] The present invention also provides a diagnostic method in which a Coriolis flowmeter, which measures the flow rate and density of a fluid flowing through a vibrating tube based on the vibration of the vibrating tube, acquires measurement parameters related to the measurement of the fluid by the Coriolis flowmeter, calculates vibrating tube parameters that indicate the state of the vibrating tube, and inputs at least one of the measurement parameters and the vibrating tube parameters into a trained machine learning model that has been trained to output a diagnosis result of the vibrating tube in response to the input of the parameters, thereby diagnosing the occurrence of an abnormal condition in the vibrating tube.

[0008] The present invention also provides a diagnostic program that causes a Coriolis flowmeter that measures the flow rate and density of a fluid flowing through a vibrating tube based on the vibration of the vibrating tube to execute a process of acquiring measurement parameters related to the measurement of the fluid by the Coriolis flowmeter, calculating vibrating tube parameters that indicate the state of the vibrating tube, and inputting at least one of the measurement parameters and the vibrating tube parameters into a trained machine learning model that has been trained to output a diagnosis result of the vibrating tube in response to the input of the parameters, thereby diagnosing the occurrence of an abnormal condition in the vibrating tube.

[0009] The present invention also provides a diagnostic system including a Coriolis flowmeter that measures the flow rate and density of a fluid flowing through a vibrating tube based on the vibration of the vibrating tube, and a diagnostic device, wherein the Coriolis flowmeter includes an acquisition unit that acquires measurement parameters related to the measurement of the fluid by the Coriolis flowmeter, and a calculation unit that calculates vibrating tube parameters that indicate a state of the vibrating tube, and the diagnostic device includes a receiving unit that receives the measurement parameters and the vibrating tube parameters, and a diagnostic unit that inputs at least one of the measurement parameters and the vibrating tube parameters into a trained machine learning model that has been trained to output a diagnosis result for the vibrating tube in response to the input of parameters, to diagnose the occurrence of an abnormal condition in the vibrating tube. Effect of the Invention

[0010] According to the present invention, it is possible to effectively diagnose the occurrence of an abnormal state in a vibrating tube of a Coriolis flowmeter. [Brief description of the drawings]

[0011] [Figure 1] 1 is a diagram illustrating a configuration example and a processing example of a Coriolis flowmeter diagnostic system according to an embodiment; [Diagram 2] 2 is a block diagram showing an example of the configuration of each device of the Coriolis flowmeter diagnostic system according to the embodiment; FIG. [Diagram 3] FIG. 4 is a diagram illustrating an acquired value storage unit of the Coriolis flowmeter according to the embodiment. [Figure 4] FIG. 4 is a diagram illustrating a calculation value storage unit of the Coriolis flowmeter according to the embodiment. [Diagram 5] FIG. 4 is a diagram illustrating a diagnosis result storage unit of the Coriolis flowmeter according to the embodiment. [Figure 6] FIG. 4 is a diagram illustrating a discrimination model storage unit of the Coriolis flowmeter according to the embodiment. [Figure 7] FIG. 4 is a diagram illustrating a regression model storage unit of the Coriolis flowmeter according to the embodiment. [Figure 8] 6 is a diagram showing a specific example 1 of a diagnosis control process of the Coriolis flowmeter according to the embodiment. FIG. [Figure 9] 11 is a diagram showing a specific example 2 of the diagnosis control process of the Coriolis flowmeter according to the embodiment. FIG. [Figure 10] 13 is a diagram showing a specific example 3 of the diagnosis control process of the Coriolis flowmeter according to the embodiment. FIG. [Figure 11] 4 is a flowchart showing an example of a flow of processing of an entire Coriolis flowmeter diagnostic system according to the embodiment. [Figure 12] 10 is a flowchart showing an example of a flow of an acquisition control process of the Coriolis flowmeter diagnostic system according to the embodiment. [Figure 13] 6 is a flowchart showing an example of a flow of a calculation control process of the Coriolis flowmeter diagnostic system according to the embodiment. [Figure 14] 5 is a flowchart showing an example of a flow of a diagnosis control process of the Coriolis flowmeter diagnosis system according to the embodiment. [Figure 15] 11A and 11B are diagrams illustrating a configuration example and a processing example of a Coriolis flowmeter diagnostic system according to a modified example of the embodiment. [Figure 16] FIG. 11 is a block diagram showing an example of the configuration of each device of a Coriolis flowmeter diagnostic system according to a modified example of the embodiment. [Figure 17] FIG. 2 is a diagram illustrating an example of a hardware configuration according to an embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0012] A Coriolis flowmeter, a diagnostic method, a diagnostic program, and a diagnostic system according to an embodiment of the present invention will be described in detail below with reference to the drawings. Note that the present invention is not limited to the embodiment described below.

[0013] [Embodiment] Below, we will explain the configuration and processing of Coriolis flowmeter diagnostic system 100 according to the embodiment, the configuration and processing of each device of Coriolis flowmeter diagnostic system 100, and the processing flow of Coriolis flowmeter diagnostic system 100, and finally the effects of the embodiment will be explained.

[0014] 1. Configuration and Processing of Coriolis Flowmeter Diagnostic System 100 A configuration and processing of a Coriolis flowmeter diagnostic system 100 according to an embodiment will be described in detail with reference to Fig. 1. Fig. 1 is a diagram showing a configuration example and processing example of the Coriolis flowmeter diagnostic system 100 according to an embodiment. Below, an example of the overall configuration of the Coriolis flowmeter diagnostic system 100, an example of the overall processing of the Coriolis flowmeter diagnostic system 100, and a basic principle of the Coriolis flowmeter diagnostic system 100 will be described in order, and finally, the effect of the Coriolis flowmeter diagnostic system 100 will be described. Note that in the embodiment, a factory production remote monitoring using a plant device that is a device installed in a plant will be described as an example, but the device and the field of use are not limited, and the present invention can also be applied to environmental measurement remote monitoring such as a power monitor, a wind power generation, a water supply and sewage monitor, and a river monitor.

[0015] (1-1. Example of the Overall Configuration of the Coriolis Flowmeter Diagnostic System 100) An example of the overall configuration of Coriolis flowmeter diagnostic system 100 will be described with reference to Fig. 1. Coriolis flowmeter diagnostic system 100 has Coriolis flowmeter 10 and operator terminal 20. Coriolis flowmeter 10 and operator terminal 20 are connected to each other via a predetermined communication network (not shown) so as to be able to communicate with each other by wire or wirelessly. The predetermined communication network may be any of various communication networks such as the Internet or a dedicated line.

[0016] (1-1-1. Coriolis flowmeter 10) Coriolis flowmeter 10 is a plant device that is installed at a plant site and measures the flow rate (mass flow rate, volumetric flow rate) and density of a fluid flowing through a piping P of the plant. Note that Coriolis flowmeter diagnostic system 100 shown in FIG. 1 may include a plurality of Coriolis flowmeters 10.

[0017] (1-1-2. Operator terminal 20) The operator terminal 20 is an information processing device used by an operator O. Note that the Coriolis flowmeter diagnostic system 100 shown in Fig. 1 may include a plurality of operator terminals 20. In addition, in the example of Fig. 1, the operator terminal 20 is realized by a desktop PC (Personal Computer), but may be realized by, for example, a smartphone, a tablet terminal, a notebook PC, a desktop PC, a mobile phone, a PDA (Personal Digital Assistant), or the like.

[0018] (1-2. Example of overall processing of Coriolis flowmeter diagnostic system 100) An example of the overall processing of Coriolis flowmeter diagnostic system 100 will be described with reference to Fig. 1. Note that the processing of steps S1 to S5 below may be executed in a different order. Also, some of the processing of steps S1 to S5 below may be omitted.

[0019] (1-2-1. Measurement parameter acquisition process) First, the Coriolis flowmeter 10 acquires measurement parameters (step S1). For example, the Coriolis flowmeter 10 acquires an angular frequency ω and a voltage ΔV corresponding to the vibration of the vibrating tube 13a (not shown) of the Coriolis flowmeter 10 from a signal detector 13c (not shown) of the Coriolis flowmeter 10. The Coriolis flowmeter 10 also acquires a frequency f, a phase difference θ, and a signal amplitude A corresponding to the vibration of the vibrating tube 13a of the Coriolis flowmeter 10 from the signal detector 13c of the Coriolis flowmeter 10.

[0020] Here, the measurement parameters are parameters related to the measurement of the fluid acquired or calculated by the Coriolis flowmeter 10, and include the angular frequency ω, voltage ΔV, frequency f, phase difference θ, signal amplitude A, etc. acquired by the Coriolis flowmeter 10, and the density ρ, mass flow rate Q, etc. of the fluid calculated by the Coriolis flowmeter 10. m , volume flow rate Q v , pressure P, flow velocity v, viscosity η, etc.

[0021] (1-2-2. Vibration tube parameter calculation process) Second, the Coriolis flowmeter 10 calculates the vibrating tube parameters (step S2). For example, the Coriolis flowmeter 10 calculates the growth rate and damping rate of the vibrating tube 13a, and calculates the mass M, spring constant K, and damping coefficient (damping constant) D of the vibrating tube 13a.

[0022] Here, the vibrating tube parameters are parameters that indicate the state of the vibrating tube 13a of the Coriolis flowmeter 10, and include the average growth rate α1, average damping rate α2, mass M, spring constant K, damping coefficient D, etc. of the vibrating tube 13a calculated by the Coriolis flowmeter 10.

[0023] (1-2-3. Corrosion diagnosis processing) Third, the Coriolis flowmeter 10 diagnoses corrosion of the vibration tube 13a (step S3). For example, the Coriolis flowmeter 10 inputs at least one of the frequency f, the phase difference θ, the signal amplitude A, the mass M, the spring constant K, and the damping coefficient D to a corrosion discrimination model DM-R, which is a trained machine learning model, to determine the occurrence of corrosion of the vibration tube 13a.

[0024] In addition, the Coriolis flowmeter 10 inputs at least one of the frequency f, phase difference θ, signal amplitude A, mass M, spring constant K, and damping coefficient D into a corrosion amount regression model RM-R, which is a trained machine learning model, to estimate the amount of corrosion of the vibrating tube 13a.

[0025] In addition, the Coriolis flowmeter 10 inputs at least one of the frequency f, phase difference θ, signal amplitude A, mass M, spring constant K, and damping coefficient D into the density error regression model RM-Δρ, which is a trained machine learning model, to estimate the density error Δρ associated with corrosion of the vibrating tube 13a.

[0026] In addition, the Coriolis flowmeter 10 inputs at least one of the frequency f, phase difference θ, signal amplitude A, mass M, spring constant K, and damping coefficient D into a flow error regression model RM-ΔQ, which is a trained machine learning model, to estimate the flow error ΔQ associated with corrosion of the vibrating tube 13a.

[0027] (1-2-4. Adhesion diagnosis processing) Fourth, the Coriolis flowmeter 10 diagnoses adhesion of the vibrating tube 13a (step S4). For example, the Coriolis flowmeter 10 inputs at least one of the frequency f, the phase difference θ, the signal amplitude A, the mass M, the spring constant K, and the damping coefficient D to an adhesion discrimination model DM-D, which is a trained machine learning model, to determine the occurrence of adhesion of the vibrating tube 13a.

[0028] In addition, the Coriolis flowmeter 10 inputs at least one of the frequency f, phase difference θ, signal amplitude A, mass M, spring constant K, and damping coefficient D into the adhesion amount regression model RM-D, which is a trained machine learning model, to estimate the adhesion amount on the vibrating tube 13a.

[0029] In addition, the Coriolis flowmeter 10 inputs at least one of the frequency f, phase difference θ, signal amplitude A, mass M, spring constant K, and damping coefficient D into the density error regression model RM-Δρ, which is a trained machine learning model, to estimate the density error Δρ associated with the adhesion of the vibrating tube 13a.

[0030] In addition, the Coriolis flowmeter 10 inputs at least one of the frequency f, phase difference θ, signal amplitude A, mass M, spring constant K, and damping coefficient D into a flow error regression model RM-ΔQ, which is a trained machine learning model, to estimate the flow error ΔQ associated with the adhesion of the vibrating tube 13a.

[0031] (1-2-5. Alarm notification processing) Fifth, Coriolis flowmeter 10 notifies operator terminal 20 of an alarm (step S5). For example, Coriolis flowmeter 10 notifies operator terminal 20 of an alarm indicating the occurrence of corrosion or adhesion of vibrating tube 13a, and prompts operator O to perform maintenance of Coriolis flowmeter 10.

[0032] Furthermore, Coriolis flowmeter 10 notifies operator terminal 20 of an alarm including the amount of corrosion or the amount of adhesion of vibrating tube 13a.

[0033] Furthermore, Coriolis flowmeter 10 notifies operator terminal 20 of an alarm including density error Δρ or flow rate error ΔQ associated with corrosion or adhesion of vibrating tube 13a.

[0034] (1-3. Basic Principles of the Coriolis Flowmeter Diagnostic System 100) Using mathematical expressions, a basic principle of Coriolis flowmeter diagnostic system 100 will be described. Below, a basic configuration of Coriolis flowmeter 10, basic processing of Coriolis flowmeter 10, and a calculation principle of parameters of Coriolis flowmeter 10 will be described in that order.

[0035] (1-3-1. Basic configuration of Coriolis flowmeter 10) The basic configuration of the Coriolis flowmeter 10 will be described. The Coriolis flowmeter 10 has an input unit 11, an output unit 12, a detection unit 13, a storage unit 14, and a control unit 15. Here, the detection unit 13 has a vibration tube 13a, an exciter 13b, and a pair of signal detectors 13c (13c-1, 13c-2). The control unit 15 has an acquisition unit 15a, a calculation unit 15b, and a diagnosis unit 15c. The acquisition unit 15a has a current amplifier 15a-1, an amplifier 15a-2, a smoothing circuit 15a-3, a control amplifier 15a-4, a voltage setting unit 15a-5, a switch 15a-6, and a multiplier 15a-7.

[0036] The vibration pipe 13a is a U-shaped pipe made of a bendable material such as metal such as stainless steel, resin, or rubber. A fluid to be measured flows inside the vibration pipe 13a. The vibrator 13b vibrates the vibration pipe 13a by a driving current supplied from a current amplifier 15a-1. The signal detector 13c is a sensor that detects the vibration of the vibration pipe 13a. The signal detector 13c outputs a voltage corresponding to the vibration of the signal detector 13c to the amplifier 15a-2.

[0037] The current amplifier 15a-1 amplifies the output signal of the multiplier 15a-7 to generate a drive current for driving the vibration tube 13a. The drive current generated by the current amplifier 15a-1 is output to the vibrator 13b of the detection unit 13. The amplifier 15a-2 amplifies and outputs the output signal of the detection unit 13 (the output signal of the signal detector 13c). The output signal of the amplifier 15a-2 is also used as a signal for detecting the Coriolis force (the amplitude and phase of the twist of the vibration tube 13a).

[0038] The smoothing circuit 15a-3 smoothes the output signal of the amplifier 15a-2 and outputs a signal indicating the amplitude of the output signal of the amplifier 15a-2. The control amplifier 15a-4 calculates the amplitude of the output signal of the smoothing circuit 15a-3 and a preset target voltage V ref and amplifies the difference between the current amplifier 15a-1 and the detection unit 13. Under the control of the calculation unit 15b, the voltage setting unit 15a-5 outputs a fixed gain setting voltage E for providing a constant gain (fixed gain) to a circuit loop consisting of the current amplifier 15a-1, the detection unit 13, the amplifier 15a-2, and the multiplier 15a-7. The value of the fixed gain setting voltage E can be changed under the control of the calculation unit 15b. Therefore, the fixed gain provided to the above circuit loop can also be changed under the control of the calculation unit 15b.

[0039] The switch 15a-6 has a first input terminal connected to the output terminal of the control amplifier 15a-4, a second input terminal connected to the output terminal of the voltage setting unit 15a-5, and one output terminal connected to the input terminal of the multiplier 15a-7, and connects either the first input terminal or the second input terminal to the output terminal under the control of the calculation unit 15b. That is, the switch 15a-6 selects either the output signal ΔV of the control amplifier 15a-4 or the fixed gain setting voltage E output from the voltage setting unit 15a-5 and outputs it to the multiplier 15a-7. Hereinafter, the signal selected by the switch 15a-6 is called "oscillation feedback current gain G (oscillation feedback gain)". The multiplier 15a-7 multiplies the output signal of the amplifier 15a-2 by the signal selected by the switch 15a-6, and outputs the multiplied signal to the current amplifier 15a-1.

[0040] The calculation unit 15b performs calculations and control necessary for diagnosing the state of the vibration tube 13a. The calculation unit 15b receives the output signal of the amplifier 15a-2, the output signal of the smoothing circuit 15a-3, and the output signal of the control amplifier 15a-4. The calculation unit 15b acquires the rate of increase or decrease of the vibration of the vibration tube 13a when the driving force with a constant oscillation feedback current gain is applied to the vibration tube 13a by the vibrator 13b. The calculation unit 15b also calculates the spring constant K and the damping coefficient D of the vibration tube 13a as mutually independent parameters based on the acquired rate of increase or decrease and the driving force applied to the vibration tube 13a. The calculation unit 15b also controls the voltage setting unit 15a-5 and the switch 15a-6. The details of the calculations and control performed by the calculation unit 15b will be described later.

[0041] (1-3-2. Basic processing of Coriolis flowmeter 10) A description will be given of basic processing of the Coriolis flowmeter 10. Below, a basic processing example 1 for measuring the flow rate of a fluid flowing inside the vibration pipe 13a and a basic processing example 2 for diagnosing an abnormal state of the vibration pipe 13a will be described.

[0042] (1-3-2-1. Basic processing example 1) When measuring the flow rate of the fluid flowing inside the vibrating tube 13a, the calculation unit 15b controls the switch 15a-6 so that the output signal ΔV of the control amplifier 15a-4 is selected by the switch 15a-6. In this way, the output signal ΔV of the control amplifier 15a-4 changes according to the output signal (amplitude) of the smoothing circuit 15a-3, and the circuit loop (a circuit loop consisting of the current amplifier 15a-1, the detection unit 13, the amplifier 15a-2, the smoothing circuit 15a-3, the control amplifier 15a-4, the switch 15a-6, and the multiplier 15a-7) converts the output signal of the smoothing circuit 15a-3 into a target voltage V ref The oscillation feedback current gain G at this time is called the oscillation feedback current gain G0. In a steady state, this circuit loop oscillates with a constant amplitude, and the vibration tube 13a vibrates with a constant amplitude. In this state, the twist of the vibration tube 13a is detected and the flow rate of the fluid flowing inside the vibration tube 13a is measured.

[0043] (1-3-2-2. Basic processing example 2) When diagnosing an abnormal state of the vibration tube 13a, the calculation unit 15b controls the switch 15a-6 so that the fixed gain setting voltage E output from the voltage setting unit 15a-5 is selected by the switch 15a-6. The calculation unit 15b also controls the voltage setting unit 15a-5 so that the fixed gain setting voltage E is set to a predetermined fixed value so that the vibration of the vibration tube 13a is attenuated or grown.

[0044] (1-3-3. Principle of Calculating the Parameters of the Coriolis Flowmeter 10) The following describes the principle of calculating the parameters of the Coriolis flowmeter 10. In the following, the oscillation feedback current gain G is sequentially switched between G1 and G2 to grow and attenuate the vibration of the vibrating tube 13a, thereby obtaining the growth rate and attenuation rate of the amplitude, and finally calculating the spring constant K and the damping coefficient D as mutually different parameters.

[0045] First, the vibration equation of the vibration tube 13a is expressed by the following formula (1): In the following formula (1), x is the vibration displacement of the vibration tube 13a, M is the mass of the entire vibration tube 13a, D is the damping coefficient, K is the spring constant of the vibration tube 13a, and G is the oscillation feedback current gain.

[0046]

number

[0047] “Gdx / dt” on the right side of the above equation (1) corresponds to the acting force based on the drive current provided to the detection unit 13 from the current amplifier 15a-1, and “G” is determined by the output signal ΔV of the control amplifier 15a-4 and the fixed gain setting voltage E.

[0048] Dividing both sides of equation (1) above by M gives equation (2) below, where D / M=2ζω, ω=sqrt(K / M), which can be transformed into equation (3) below and equation (4) below, which gives equation (5) below. In equations (3) to (5) below, ω is the angular frequency.

[0049]

number

[0050]

number

[0051]

number

[0052]

number

[0053] As described above, the vibration displacement x of the vibration tube 13a has a waveform of a sine wave that attenuates or grows. The output signal of the detection unit 13, which is proportional to the vibration displacement x of the vibration tube 13a, also has a waveform of a sine wave that attenuates or grows.

[0054] When the oscillation feedback current gain G is set to G1 (G1>G0), the rate of increase / decrease of the integral value of the squared value output by the detection unit 13 is set to α1 (>1), and when the oscillation feedback current gain G is set to G2 (G0>G2), the rate of increase / decrease of the integral value of the squared value output by the detection unit 13 is set to α2 (<1), as shown in the following equations (6) and (7).

[0055]

number

[0056]

number

[0057] From the above formulas (6) and (7), we obtain the following formula (8): Note that a rate of increase or decrease > 1 indicates a growth in amplitude, and a rate of increase or decrease < 1 indicates a decay in amplitude.

[0058]

number

[0059] By using the above procedure, the spring constant K expressed by the following formula (9) and the damping coefficient D expressed by the following formula (10) can be calculated.

[0060]

number

[0061]

number

[0062] (1-3-3. Details of the Calculation Process of the Parameters of the Coriolis Flowmeter 10) A detailed description will be given of the calculation process of the parameters of the Coriolis flowmeter 10. The following describes the increase / decrease rate calculation process, the switch changeover process, and the vibration pipe parameter calculation process in that order.

[0063] (1-3-3-1. Increase / decrease rate calculation process) The increase / decrease rate calculation process is a process for calculating the increase / decrease rate of the vibration of the vibration tube 13a when the vibrator 13b applies a driving force with a constant oscillation feedback current gain to the vibration tube 13a. The increase / decrease rate calculation process is executed by the acquisition unit 15a and the calculation unit 15b. It is assumed that, at the time when the increase / decrease rate calculation process is started, the output signal ΔV of the control amplifier 15a-4 is selected by the switch 15a-6, and the above-mentioned basic processing example 1 is being executed.

[0064] First, the acquiring unit 15a acquires the angular frequency ω in the basic processing example 1. Here, the calculating unit 15b can acquire the angular frequency ω based on the output signal (Asin(ωt)) of the amplifier 15a-2 and the like.

[0065] Secondly, the acquiring unit 15a acquires the output signal ΔV of the control amplifier 15a-4. By the above-mentioned second process, the oscillation feedback current gain G in the basic process example 1 can be obtained.

[0066] Third, the calculation unit 15b judges whether the diagnosis start time t0 has been reached. If the calculation unit 15b judges that the diagnosis start time t0 has not been reached, the calculation unit 15b returns to the first process of acquiring the angular frequency ω. On the other hand, if the calculation unit 15b judges that the diagnosis start time t0 has been reached, the calculation unit 15b proceeds to the fourth process described below.

[0067] Fourth, calculation unit 15b switches switch 15a-6 from the side that selects output signal ΔV of control amplifier 15a-4 to the side that selects fixed gain setting voltage E, which is the output value of voltage setting unit 15a-5. Furthermore, calculation unit 15b sets the value of fixed gain setting voltage E of voltage setting unit 15a-5 to a value that makes oscillation feedback current gain G G1. By the above fourth process, the increase / decrease rate α of the integral value of the squared value output by detection unit 13 is set to α1 (>1).

[0068] Fifth, the calculation unit 15b calculates the rate of increase or decrease (growth rate) of the amplitude of the sine wave for each period i=1 to N (i, N: integers) until time t1, i.e., for L seconds. As expressed by the following formula (11), the calculation unit 15b calculates the sum of squares Sj of the sine wave: A×exp(-ζωt+G1 / 2M)t×sin(ωt+φ) for each period (i=1 to N) of Δt.

[0069]

number

[0070] If the sampling rate output by the detection unit 13 is 0.001 seconds and the period is Δt, that is, the number of samples N is 1000, then the integration time of Sj is 1 second. If this is repeated L times, the total time is L seconds. The calculation unit 15b calculates the rate of increase or decrease R of adjacent amplitudes according to the following formula (12), and repeats the same process from j=1 to L.

[0071]

number

[0072] Sixth, the calculation unit 15b calculates an average value R of the rate of increase or decrease R calculated L times as shown in the following formula (13). AVE Calculate.

[0073]

number

[0074] The calculation unit 15b calculates the rate of increase / decrease α1 (>1) as an average value R per unit time from the above formula (13) as shown in the following formula (14). AVE It is calculated as follows.

[0075]

number

[0076] Seventh, at time t1, calculation unit 15b sets the value of fixed gain setting voltage E of voltage setting unit 15a-5 to a value at which oscillation feedback current gain G becomes G2. By the seventh process described above, the increase / decrease rate α of the integral value of the squared value output by detection unit 13 is set to α2 (<1). Also, by the seventh process described above, the increase / decrease rate α is switched from α1 (>1) to α2 (<1) at time t1. Also, the amplitude of the output signal (Asin(ωt)) of amplifier 15a-2, which had increased until time t1, starts to decrease from time t1.

[0077] Eighth, the calculation unit 15b calculates the rate of increase or decrease (attenuation rate) of the amplitude of the sine wave for each of the periods i=1 to N until time t2, i.e., for L seconds. The eighth process is executed in the same manner as the fifth process. That is, the calculation unit 15b calculates the sum of squares Sj of the sine wave: A×exp(-ζωt+G2 / 2M)t×sin(ωt+φ) for each of the periods (i=1 to N) of Δt.

[0078] Ninthly, the calculation unit 15b calculates an average value R of the rate of increase or decrease R calculated L times. AVEThe ninth process is executed in the same manner as the sixth process. That is, the calculation unit 15b calculates the rate of increase / decrease α2 (<1) as the average value R per unit time from the above formula (13) as shown in the following formula (15). AVE It is calculated as follows.

[0079]

number

[0080] (1-3-3-2. Switching process) The switch changeover process is a process for executing changeover of the switch 15a-6. The switch changeover process is executed by the calculation unit 15b.

[0081] At time t2, calculation unit 15b switches switch 15a-6 from the side that selects fixed gain setting voltage E, which is the output value of voltage setting unit 15a-5, to the side that selects output signal ΔV of control amplifier 15a-4. The above switch changeover process returns to the above-mentioned basic processing example 1, so that the increase / decrease rate α becomes approximately 1, and the amplitude of the sine wave is maintained at an approximately constant value.

[0082] (1-3-3-3. Vibration tube parameter calculation process) The vibration pipe parameter calculation process is a process for calculating the mass M, spring constant K, and damping coefficient D of the vibration pipe 13a as parameters based on the increase / decrease rate calculated by the increase / decrease rate calculation process described above and the driving force applied to the vibration pipe 13a. The vibration pipe parameter calculation process is executed by the calculation unit 15b.

[0083] The calculation unit 15b calculates the mass M of the vibration tube 13a using the following formula (16) in which ζω is eliminated from the above formulas (14) and (15).

[0084]

number

[0085] Moreover, the calculation unit 15b calculates the spring constant K and the damping coefficient D by using the above formulas (9) and (10).

[0086] (1-4. Effects of the Coriolis flowmeter diagnostic system 100) In the following, an overview and problems of Coriolis flowmeter diagnostic systems 100P (100P-1, 100P-2, 100P-3, 100P-4) according to Reference Techniques 1 to 4 will be described, and then the effects of Coriolis flowmeter diagnostic system 100 will be described.

[0087] (1-4-1. Overview of Coriolis Flowmeter Diagnostic System 100P) An overview of the Coriolis flowmeter diagnostic systems 100P according to Reference Technologies 1 to 4 will be described below.

[0088] The Coriolis flowmeter diagnostic system 100P-1 according to the reference technique 1 provides a measurement system capable of identifying a plurality of parameters related to the vibration (resonance) of the vibrating tube 13a of the Coriolis flowmeter 10. The Coriolis flowmeter diagnostic system 100P-1 can use the parameters to diagnose corrosion or adhesion of the vibrating tube 13a, perform maintenance, etc. In the Coriolis flowmeter diagnostic system 100P-1, the parameters related to the vibration of the Coriolis flowmeter 10 (such as the spring constant and damping coefficient of the vibrating tube 13a) can be identified by evaluating the relationship between the excitation signal F(t) of the vibration amplitude and phase and the corresponding response signal y(t) in the closed loop control system of the Coriolis flowmeter 10 using a mathematical model. At this time, the Coriolis flowmeter diagnostic system 100P-1 provides an excitation signal such as -45°, 0°, and 45° in the phase of the vibration, and identifies the parameters using a mathematical model for the relationship with the corresponding response signal y(t).

[0089] The Coriolis flowmeter diagnostic system 100P-2 according to the reference technique 2 corrects density measurement errors caused by changes in temperature, pressure, and fluid composition of the Coriolis flowmeter 10. In density measurement using the Coriolis flowmeter 10, density fluctuations may occur over time in an application. In this case, if the density fluctuations increase over time, temperature changes in the process or corrosion or adhesion of the vibrating tube 13a is suspected. In the Coriolis flowmeter diagnostic system 100P-2, regarding temperature changes, the fluid density is measured at a preset reference temperature, and density errors caused by temperature changes are corrected. In addition, the Coriolis flowmeter diagnostic system 100P-2 can also perform corrections for pressure and fluid composition. Here, in the Coriolis flowmeter diagnostic system 100P-2, if the corrected density measurement value deviates from the threshold value, the Coriolis flowmeter diagnostic system 100P-2 notifies the operator O of the possibility that corrosion or adhesion of the vibrating tube 13a has occurred.

[0090] The Coriolis flowmeter diagnostic system 100P-3 according to Reference Technology 3 and the Coriolis flowmeter diagnostic system 100P-4 according to Reference Technology 4 diagnose the corrosion and adhesion of the vibration tube 13a of the Coriolis flowmeter 10 and notify the operator O of the maintenance timing. In the Coriolis flowmeter diagnostic system 100P-3 and the Coriolis flowmeter diagnostic system 100P-4, a voltage setting unit that generates a fixed gain setting voltage E is added to the control circuit of the Coriolis flowmeter 10, and when the corrosion diagnosis or adhesion diagnosis of the vibration tube 13a is performed, the fixed gain of the vibration tube 13a is changed to attenuate or grow the vibration amplitude of the vibration tube 13a, and the spring constant and damping coefficient of the vibration tube 13a are measured from the inclination. In the Coriolis flowmeter diagnostic system 100P-3 and the Coriolis flowmeter diagnostic system 100P-4, since the spring constant is an index of the progress of corrosion and the damping coefficient is an index of the progress of adhesion, an alarm is notified when the threshold value set by the operator O is exceeded. Furthermore, Coriolis flowmeter diagnostic system 100P-3 and Coriolis flowmeter diagnostic system 100P-4 can predict the time when the calculated spring constant and damping coefficient will exceed the threshold value, and notify operator O of the timing for maintenance.

[0091] (1-4-2. Problems with the Coriolis flowmeter diagnostic system 100P) Problems with the Coriolis flowmeter diagnostic systems 100P according to the reference techniques 1 to 4 will be described below.

[0092] Although the Coriolis flowmeter diagnostic system 100P-1 according to the reference technique 1 can calculate the spring constant and damping coefficient of the vibrating tube 13a, it is difficult to notify the operator O of the occurrence of corrosion or adhesion in the application. In addition, since the Coriolis flowmeter diagnostic system 100P-1 cannot estimate the degree of corrosion or adhesion, it is difficult for the operator O to determine whether or not maintenance of the Coriolis flowmeter 10 is required.

[0093] In the Coriolis flowmeter diagnostic system 100P-2 according to the reference technique 2, it is possible to diagnose the possibility of corrosion or adhesion of the vibrating tube 13a when the density deviates significantly from the expected value, but it is difficult to notify the operator O of the occurrence of corrosion or adhesion in the application. In addition, since the Coriolis flowmeter diagnostic system 100P-2 cannot estimate the degree of corrosion or adhesion, it is difficult for the operator O to determine whether or not maintenance of the Coriolis flowmeter 10 is required.

[0094] In the Coriolis flowmeter diagnostic system 100P-3 according to Reference Technology 3 and the Coriolis flowmeter diagnostic system 100P-4 according to Reference Technology 4, changes in the mechanical properties of the vibrating tube 13a are used as diagnostic results, such as the spring constant K for corrosion and the damping coefficient D for adhesion, but it is difficult for the operator O to quantitatively determine the progress of corrosion or adhesion from the diagnostic results. Therefore, the operator O must independently set the conditions for alarm notification and the timing of maintenance based on intuition and experience.

[0095] (1-4-3. Overview of Coriolis flowmeter diagnostic system 100) The Coriolis flowmeter diagnostic system 100 according to the embodiment executes the following processes. First, the Coriolis flowmeter 10 acquires measurement parameters such as frequency f, phase difference θ, and signal amplitude A. Second, the Coriolis flowmeter 10 calculates vibration tube parameters D such as mass M, spring constant K, and damping coefficient of the vibration tube 13a. Third, the Coriolis flowmeter 10 inputs parameters to the learned corrosion discrimination model DM-R and corrosion amount regression model RM-R to diagnose the occurrence and amount of corrosion of the vibration tube 13a. Diagnoses corrosion of the vibration tube 13a. Fourth, the Coriolis flowmeter 10 inputs parameters to the learned adhesion discrimination model DM-D and corrosion amount regression model RM-D to diagnose the occurrence and amount of adhesion of the vibration tube 13a. Fifth, when the Coriolis flowmeter 10 diagnoses corrosion or adhesion, it notifies an alarm to the operator terminal 20.

[0096] (1-4-4. Effects of the Coriolis flowmeter diagnostic system 100) As described above, Coriolis flowmeter diagnostic system 100 quantitatively estimates the state of corrosion or adhesion and notifies operator O, rather than issuing an alarm for corrosion or adhesion or providing information such as spring constant K and damping coefficient D. Coriolis flowmeter diagnostic system 100 also estimates density error and flow rate error from the amount of corrosion and adhesion, corrects them, and notifies operator O of the estimated error, making it easier for operator O to determine the timing of maintenance or replacement of Coriolis flowmeter 10. That is, Coriolis flowmeter diagnostic system 100 effectively diagnoses the occurrence of an abnormal state in vibrating tube 13a of Coriolis flowmeter 10, and can provide operator O with more useful information about the diagnosis results for corrosion and adhesion of vibrating tube 13a.

[0097] 2. Configuration and Processing of Each Device of Coriolis Flowmeter Diagnostic System 100 The configuration and processing of each device included in the Coriolis flowmeter diagnostic system 100 shown in Fig. 1 will be described with reference to Fig. 2. Fig. 2 is a block diagram showing an example of the configuration of each device of the Coriolis flowmeter diagnostic system 100 according to the embodiment. Below, an example of the configuration of the entire Coriolis flowmeter diagnostic system 100 according to the embodiment will be described, and then an example of the configuration and processing of the Coriolis flowmeter 10, an example of the configuration and processing of the operator terminal 20, and an example of the configuration and processing of the external sensor 30 according to the embodiment will be described in detail.

[0098] (2-1. Example of the overall configuration of the Coriolis flowmeter diagnostic system 100) An example of the overall configuration of Coriolis flowmeter diagnostic system 100 shown in FIG. 1 will be described with reference to FIG. 2. As shown in FIG. 2, Coriolis flowmeter diagnostic system 100 has Coriolis flowmeter 10, operator terminal 20, and external sensor 30. Coriolis flowmeter 10 is communicatively connected to operator terminal 20 and external sensor 30 via a communication network N realized by the Internet, a dedicated line, or the like. Coriolis flowmeter 10 is installed in a plant, but may be installed in a cloud environment, an on-premise environment, an edge environment, or the like. Furthermore, external sensor 30 is installed in the plant.

[0099] (2-2. Configuration and Processing Examples of Coriolis Flowmeter 10) A configuration example and a processing example of the Coriolis flowmeter 10 will be described with reference to Fig. 2. The Coriolis flowmeter 10 is a measuring device that measures the flow rate and density of a fluid flowing through a vibrating tube 13a based on the Coriolis force acting on the vibrated vibrating tube 13a, and has an input unit 11, an output unit 12, a detection unit 13, a storage unit 14, and a control unit 15.

[0100] (2-2-1. Input section 11) The input unit 11 controls input of various information to the Coriolis flowmeter 10. For example, the input unit 11 is realized by a mouse, a keyboard, or the like, and accepts input of various information to the Coriolis flowmeter 10.

[0101] (2-2-2. Output section 12) Output unit 12 controls the output of various information from Coriolis flowmeter 10. For example, output unit 12 is realized by a display or the like, and displays various information stored in Coriolis flowmeter 10.

[0102] (2-2-3. Detection unit 13) The detection unit 13 detects various types of information and includes a vibration tube 13a, a vibrator 13b, and a signal detector 13c.

[0103] (2-2-3-1. Vibration tube 13a) The vibration pipe 13a vibrates at a constant amplitude and frequency and is realized by a pair of U-shaped pipes made of a bendable material such as stainless steel metal, resin, or rubber.

[0104] (2-2-3-2. Vibrator 13b) The vibrator 13b vibrates the vibration tube 13a by a driving current supplied from an acquisition unit 15a of the control unit 15, which will be described later.

[0105] (2-2-3-3. Signal detector 13c) The signal detector 13c acquires various signals. For example, the vibration sensor 13c-A, which is a signal detector 13c that detects vibration, detects the vibration of the vibrating tube 13a and acquires a signal corresponding to the vibration. The temperature sensor 13c-B, which is a signal detector 13c that detects temperature, detects the temperature of the fluid flowing through the vibrating tube 13a and acquires a signal corresponding to the temperature. The pressure sensor 13c-C, which is a signal detector 13c that detects pressure, detects the pressure of the fluid flowing through the vibrating tube 13a and acquires a signal corresponding to the pressure.

[0106] (2-2-4. Storage section 14) The storage unit 14 stores various information referenced when the control unit 15 operates and various information acquired when the control unit 15 operates. The storage unit 14 has an acquired value storage unit 14a, a calculated value storage unit 14b, a diagnosis result storage unit 14c, a discrimination model storage unit 14d, and a regression model storage unit 14e. Here, the storage unit 14 can be realized by, for example, a semiconductor memory element such as a random access memory (RAM) or a flash memory, or a storage device such as a hard disk or an optical disk. Note that, in the example of FIG. 2, the storage unit 14 is installed inside the Coriolis flowmeter 10, but may be installed outside the Coriolis flowmeter 10, or a plurality of storage units may be installed.

[0107] (2-2-4-1. Acquired value storage unit 14a) The acquired value storage unit 14a stores the acquired values. For example, the acquired value storage unit 14a stores measurement parameters acquired by an acquisition unit 15a of the control unit 15 described later. Here, an example of data stored in the acquired value storage unit 14a will be described with reference to FIG. 3. FIG. 3 is a diagram showing an example of the acquired value storage unit 14a of the Coriolis flowmeter 10 according to the embodiment. In the example of FIG. 3, the acquired value storage unit 14a has items such as "time", "angular frequency", "voltage", "frequency", "phase difference", and "signal amplitude".

[0108] "Time" indicates the time when the Coriolis flowmeter 10 acquired the measurement parameters, and is expressed, for example, in terms of year, month, day, hour, minute, and second. "Angular frequency" indicates the angular frequency ω of the output signal of the vibrating tube 13a acquired by the Coriolis flowmeter 10 from the signal detector 13c, and is expressed, for example, in units of rad / s (radians per second). "Voltage" indicates the voltage ΔV of the output signal of the vibrating tube 13a acquired by the Coriolis flowmeter 10 from the signal detector 13c, and is expressed, for example, in units of V (volts). "Frequency" indicates the frequency f of the output signal of the vibrating tube 13a acquired by the Coriolis flowmeter 10 from the signal detector 13c, and is expressed, for example, in units of Hz (Hertz). "Phase difference" indicates the phase difference θ of the output signal of the vibrating tube 13a acquired by the Coriolis flowmeter 10 from the signal detector 13c, and is expressed, for example, in units of s (seconds). "Signal amplitude" indicates the signal amplitude A of the output signal of the vibrating tube 13a acquired by the Coriolis flowmeter 10 from the signal detector 13c, and is expressed in units of m (meters), for example.

[0109] That is, FIG. 3 shows an example in which data having the acquired values ​​of {time: "time T001", angular frequency: "angular frequency ω001", voltage: "voltage ΔV001", frequency: "frequency f001", phase difference: "phase difference θ001", signal amplitude: "signal amplitude A001"}, {time: "time T002", angular frequency: "angular frequency ω002", voltage: "voltage ΔV002", frequency: "frequency f002", phase difference: "phase difference θ002", signal amplitude: "signal amplitude A002"}, {time: "time T003", angular frequency: "angular frequency ω003", voltage: "voltage ΔV003", frequency: "frequency f003", phase difference: "phase difference θ003", signal amplitude: "signal amplitude A003"} are stored in the acquired value storage unit 14a.

[0110] (2-2-4-2. Calculated value storage unit 14b) The calculated value storage unit 14b stores the calculated values. For example, the calculated value storage unit 14b stores the density ρ of the flow rate, the mass flow rate Q of the flow rate calculated by the calculation unit 15b of the control unit 15 described later. m , and the volumetric flow rate Q vIn addition, as vibration tube parameters indicating the state of the vibration tube 13a, the growth rate average value α1, the damping rate average value α2, the mass M, the spring constant K, and the damping coefficient D of the vibration tube 13a are stored.

[0111] An example of data stored in the calculated value storage unit 14b will now be described with reference to Fig. 4. Fig. 4 is a diagram showing an example of the calculated value storage unit 14b of the Coriolis flowmeter 10 according to the embodiment. In the example of Fig. 4, the calculated value storage unit 14b has items such as "density", "mass flow rate", "volume flow rate", "average growth rate", "average damping rate", "mass", "spring constant", and "damping coefficient".

[0112] "Time" indicates the time when the Coriolis flowmeter 10 acquired the measurement parameters, and is expressed, for example, in terms of year, month, day, hour, minute, and second. "Density" indicates the density ρ of the fluid calculated by the Coriolis flowmeter 10, and is expressed, for example, in units of kg / m 3 The "mass flow rate" is the mass flow rate Q of the fluid calculated by the Coriolis flowmeter 10. m The "volumetric flow rate" is the mass flow rate Q of the fluid calculated by the Coriolis flowmeter 10. m For example, the unit m 3 / s. "Average growth rate" indicates the average growth rate of the vibrating tube 13a calculated by the Coriolis flowmeter 10. "Average damping rate" indicates the average damping rate of the vibrating tube 13a calculated by the Coriolis flowmeter 10. "Mass" indicates the total mass M of the vibrating tube 13a calculated by the Coriolis flowmeter 10 and is expressed in units of g, for example. "Spring constant" indicates the spring constant K of the vibrating tube 13a calculated by the Coriolis flowmeter 10. "Damping coefficient" indicates the damping coefficient D of the vibrating tube 13a calculated by the Coriolis flowmeter 10.

[0113] That is, FIG. 4 shows an example in which the calculated values ​​stored in calculated value storage unit 14b are data such as {time: "time T001", density: "density ρ001", mass flow rate: "mass flow rate Qm001", volumetric flow rate: "volume flow rate Qv001", average growth rate: "average growth rate α1-001", average damping rate: "average damping rate α2-001", mass: "mass M-001", spring constant: "spring constant K001", damping coefficient: "damping coefficient D001"}.

[0114] (2-2-4-3. Diagnosis result storage unit 14c) The diagnosis result storage unit 14c stores the diagnosis result. For example, the diagnosis result storage unit 14c stores the diagnosis result output by a diagnosis unit 15c of the control unit 15 described later. Here, an example of data stored in the diagnosis result storage unit 14c will be described with reference to FIG. 5. FIG. 5 is a diagram showing an example of the diagnosis result storage unit 14c of the Coriolis flowmeter 10 according to the embodiment. In the example of FIG. 5, the diagnosis result storage unit 14c has items such as "time", "corrosion", "amount of corrosion", "adhesion", "amount of adhesion", "density error", and "flow rate error".

[0115] "Time" indicates the time when the Coriolis flowmeter 10 acquired the measurement parameters, and is expressed, for example, in units of year, month, day, hour, minute, and second. "Corrosion" indicates the presence or absence of corrosion of the vibration tube 13a diagnosed by the diagnosis unit 15c, and is expressed, for example, as "○" if corrosion occurs, and as "-" if no corrosion occurs. "Amount of corrosion" indicates the amount of corrosion of the vibration tube 13a diagnosed by the diagnosis unit 15c, and is expressed, for example, in units of g, mm, %, etc. "Adhesion" indicates the presence or absence of adhesion of the vibration tube 13a diagnosed by the diagnosis unit 15c, and is expressed, for example, in units of g, mm, %, etc. "Amount of adhesion" indicates the amount of adhesion of the vibration tube 13a diagnosed by the diagnosis unit 15c, and is expressed, for example, in units of g, mm, %, etc. "Density error" indicates the density error Δρ of the fluid due to corrosion or adhesion diagnosed by the diagnosis unit 15c, and is expressed, for example, in units of kg / m 3 The "flow rate error" is the mass flow rate error ΔQ of the fluid due to corrosion or adhesion diagnosed by the diagnosis unit 15c. m and volume flow error ΔQv For example, the units kg / s and m 3 It is expressed as / s.

[0116] That is, FIG. 5 shows an example in which the diagnosis result data, which is {time: "Time T001", corrosion: "-", amount of corrosion: "-", adhesion: "○", amount of adhesion: "Xmm", density error: "Δρ", flow rate error: "ΔQ"}... is stored in the diagnosis result memory unit 14c.

[0117] (2-2-4-4. Discriminant model storage unit 14d) The discrimination model storage unit 14d stores the discrimination model DM. For example, the discrimination model storage unit 14d stores the discrimination model DM (corrosion discrimination model DM-R, adhesion discrimination model DM-D) used by the diagnosis unit 15c of the control unit 15 described later. Here, an example of data stored in the discrimination model storage unit 14d will be described with reference to FIG. 6. FIG. 6 is a diagram showing an example of the discrimination model storage unit 14d of the Coriolis flowmeter 10 according to the embodiment. In the example of FIG. 6, the discrimination model storage unit 14d has an item such as "discrimination model."

[0118] A "discrimination model" is model data of a trained machine learning model, and is data including, for example, execution data for executing the algorithm of the discriminant model DM, model parameters, hyperparameters, etc.

[0119] That is, Figure 6 shows an example in which data such as "corrosion discrimination model DM-R001", "adhesion discrimination model DM-D001", ... are stored in the discrimination model memory unit 14d as model data of the discrimination model DM used by the diagnosis unit 15c.

[0120] (2-2-4-5. Regression model storage unit 14e) The regression model storage unit 14e stores the regression model RM. For example, the regression model storage unit 14e stores the regression models RM (corrosion amount regression model RM-R, adhesion amount regression model RM-D, density error regression model RM-Δρ, and flow rate error regression model RM-ΔQ) used by a diagnosis unit 15c of the control unit 15 described later. Here, an example of data stored in the regression model storage unit 14e will be described with reference to FIG. 7. FIG. 7 is a diagram showing an example of the regression model storage unit 14e of the Coriolis flowmeter 10 according to the embodiment. In the example of FIG. 7, the regression model storage unit 14e has an item such as "regression model."

[0121] A "regression model" is model data of a trained machine learning model, and is data including, for example, execution data for executing the algorithm of the regression model RM, model parameters, hyperparameters, etc.

[0122] That is, FIG. 7 shows an example in which the following data are stored in the regression model memory unit 14e as model data for the regression model RM used by the diagnosis unit 15c: “corrosion amount regression model RM-R001,” “adhesion amount regression model RM-D001,” “density error regression model RM-Δρ001,” “flow rate error regression model RM-ΔQ001,” ...

[0123] (2-2-5. Control unit 15) Control unit 15 is responsible for overall control of Coriolis flowmeter 10. Control unit 15 has acquisition unit 15a, calculation unit 15b, and diagnosis unit 15c. Here, control unit 15 can be realized by, for example, an electronic circuit such as a central processing unit (CPU) or a micro processing unit (MPU), or an integrated circuit such as an application specific integrated circuit (ASIC) or a field programmable gate array (FPGA).

[0124] (2-2-5-1. Acquisition part 15a) The acquiring unit 15a acquires various information. The acquiring unit 15a stores the acquired various information in the storage unit 14. For example, the acquiring unit 15a acquires measurement parameters related to the measurement of the fluid of the Coriolis flowmeter 10. At this time, the acquiring unit 15a acquires the frequency f, phase difference θ, and signal amplitude A based on the vibration of the vibration tube 13a as the measurement parameters. The acquiring unit 15a may also acquire the angular frequency ω, voltage ΔV, etc. based on the vibration of the vibration tube 13a via the signal detector 13c as the measurement parameters. The acquiring unit 15a may also acquire the pressure P of the fluid, etc. via the external sensor 30 as the measurement parameters. The acquiring unit 15a stores the acquired measurement parameters in the acquired value storage unit 14a.

[0125] To explain a specific example, the acquisition unit 15a acquires the following as measurement parameters: {time: "Time T001", angular frequency: "Angular frequency ω001", voltage: "Voltage ΔV001", frequency: "Frequency f001", phase difference: "Phase difference θ001", signal amplitude: "Signal amplitude A001"}, {time: "Time T002", angular frequency: "Angular frequency ω002", voltage: "Voltage ΔV002", frequency: "Frequency f002", phase difference: "Phase difference θ002", signal amplitude: "Signal amplitude A002"}, {time: "Time T003", angular frequency: "Angular frequency ω003", voltage: "Voltage ΔV003", frequency: "Frequency f003", phase difference: "Phase difference θ003", signal amplitude: "Signal amplitude A003"}, and stores them in the acquisition value storage unit 14a.

[0126] (2-2-5-2. Calculation section 15b) The calculation unit 15b calculates various pieces of information. The calculation unit 15b stores the calculated various pieces of information in the storage unit 14. The vibration pipe parameter calculation process and the parameter change amount calculation process will be described below.

[0127] (Vibration tube parameter calculation process) The calculation unit 15b executes a vibration pipe parameter calculation process. For example, the calculation unit 15b calculates vibration pipe parameters that indicate the state of the vibration pipe 13a. At this time, the calculation unit 15b calculates the mass M, spring constant K, and damping coefficient D of the vibration pipe 13a as the vibration pipe parameters. The calculation unit 15b may also calculate the growth rate average value α1 and damping rate average value α2 of the vibration pipe 13a as the vibration pipe parameters.

[0128] A specific example of the vibration tube parameter calculation process will be described. The calculation unit 15b calculates the "growth rate average value α1-001" of the vibration tube 13a based on the oscillation feedback current gain G1, and calculates the "decay rate average value α2-001" based on the oscillation feedback current gain G2. The calculation unit 15b also calculates the "mass M001" which is the mass M of the vibration tube 13a at "time T001" by substituting the oscillation feedback current gain G1, the oscillation feedback current gain G2, the "growth rate average value α1-001", and the "decay rate average value α2-001" into the formula (8) above, and stores it in the calculated value storage unit 14b. The calculation unit 15b also calculates the "spring constant K001" which is the spring constant K of the vibration tube 13a at "time T001" by substituting the "mass M001" and the "angular frequency ω001" into the formula (9) above, and stores it in the calculated value storage unit 14b. In addition, the calculation unit 15b calculates the "damping coefficient D001", which is the damping coefficient D of the vibration tube 13a at "time T001", by substituting the oscillation feedback current gain G1, the "growth rate average value α1-001", the "mass M001", and the period Δt into the above formula (10), and stores the calculated value in the calculation value memory unit 14b.

[0129] (Parameter change calculation process) The calculation unit 15b executes a parameter change amount calculation process. For example, the calculation unit 15b calculates the change rate of the parameter for a predetermined period. At this time, the calculation unit 15b calculates the change rates of the frequency f, the phase difference θ, the signal amplitude A, the mass M, the spring constant K, and the damping coefficient D for the predetermined period.

[0130] To explain a specific example of the parameter change calculation process, the calculation unit 15b calculates the average change rate ΔM of "mass M001", "mass M002", "mass M003", ..., "mass M060" calculated respectively at "time T001", "time T002", "time T003", ..., "time T060".

[0131] (2-2-5-3. Diagnosis unit 15c) The diagnosis unit 15c executes various diagnoses. The diagnosis unit 15c stores various information in the storage unit 14. The diagnosis unit 15c also acquires various information stored in the storage unit 14. The diagnosis unit 15c inputs at least one of the measurement parameters and the vibration pipe parameters to a trained machine learning model that has been trained to output a diagnosis result for the vibration pipe 13a in response to the input of the parameters, and diagnoses the occurrence of an abnormal state in the vibration pipe 13a. The corrosion diagnosis process, the adhesion diagnosis process, and the alarm notification process are described below.

[0132] (Corrosion diagnostic treatment) The diagnosis unit 15c executes a corrosion diagnosis process. For example, the diagnosis unit 15c inputs at least one of the frequency f, phase difference θ, signal amplitude A, mass M, spring constant K, and damping coefficient D to the corrosion discrimination model DM-R as a trained machine learning model, and diagnoses the occurrence of corrosion in the vibration tube 13a as an abnormal state in the vibration tube 13a. The diagnosis unit 15c also inputs at least one of the frequency f, phase difference θ, signal amplitude A, mass M, spring constant K, and damping coefficient D to the corrosion amount regression model RM-R as a trained machine learning model, and estimates the amount of corrosion in the vibration tube 13a.

[0133] The diagnosis unit 15c inputs at least one of frequency f, phase difference θ, signal amplitude A, mass M, spring constant K, and damping coefficient D to a density error regression model RM-Δρ as a trained machine learning model, and estimates a density error Δρ of the fluid due to the occurrence of corrosion. The diagnosis unit 15c also inputs at least one of frequency f, phase difference θ, signal amplitude A, mass M, spring constant K, and damping coefficient D to a flow rate error regression model RM-ΔQ as a trained machine learning model, and estimates a flow rate error ΔQ of the fluid due to the occurrence of corrosion.

[0134] The diagnoser 15c inputs the calculated parameter change rate into the corrosion discrimination model DM-R as a trained machine learning model, and diagnoses the occurrence of corrosion of the vibration tube 13a as an abnormal state in the vibration tube 13a.

[0135] To explain a specific example of the corrosion diagnosis processing, the diagnosis unit 15c inputs "frequency f005," "phase difference θ005," "signal amplitude A005," "mass M005," "spring constant K005," and "damping coefficient D005" into the corrosion discrimination model DM-R, obtains "corrosion present" output from the corrosion discrimination model DM-R in response to the input, and diagnoses that corrosion has occurred in the vibration tube 13a at "time T005."

[0136] In addition, the diagnosis unit 15c inputs the "frequency f005," "phase difference θ005," "signal amplitude A005," "mass M005," "spring constant K005," and "damping coefficient D005" into the corrosion amount regression model RM-R, acquires "Yg" output by the corrosion amount regression model RM-R in accordance with the input, and estimates that corrosion of the corrosion amount Y [g] has occurred in the vibrating tube 13a at "time T005."

[0137] In addition, the diagnosis unit 15c inputs the “frequency f005,” “phase difference θ005,” “signal amplitude A005,” “mass M005,” “spring constant K005,” and “damping coefficient D005” into the density error regression model RM-Δρ, acquires “Δρ-R005” output by the density error regression model RM-Δρ in accordance with the input, and estimates the error in density ρ of “Δρ-R005” associated with the corrosion of the vibrating tube 13a at “time T005.”

[0138] In addition, the diagnostic unit 15c inputs the "frequency f005," "phase difference θ005," "signal amplitude A005," "mass M005," "spring constant K005," and "damping coefficient D005" into the flow error regression model RM-ΔQ, acquires "ΔQm-R005" output by the flow error regression model RM-ΔQ in accordance with the input, and estimates the error in the mass flow rate Qm of "ΔQm-R005" due to corrosion of the vibrating tube 13a at "time T005."

[0139] In addition, the diagnosis unit 15c inputs "mass M005" and the average change rate ΔM calculated by the calculation unit 15b into the corrosion discrimination model DM-R, obtains "corrosion present" output by the corrosion discrimination model DM-R in response to the input, and diagnoses that corrosion has occurred in the vibrating tube 13a at "time T005".

[0140] (Adhesion diagnosis processing) The diagnosis unit 15c executes adhesion diagnosis processing. For example, the diagnosis unit 15c inputs at least one of the frequency f, phase difference θ, signal amplitude A, mass M, spring constant K, and damping coefficient D to the adhesion discrimination model DM-D as a trained machine learning model, and diagnoses the occurrence of adhesion on the vibration tube 13a as an abnormal state in the vibration tube 13a. The diagnosis unit 15c also inputs at least one of the frequency f, phase difference θ, signal amplitude A, mass M, spring constant K, and damping coefficient D to the adhesion amount regression model RM-D as a trained machine learning model, and estimates the adhesion amount on the vibration tube 13a.

[0141] The diagnosis unit 15c inputs at least one of frequency f, phase difference θ, signal amplitude A, mass M, spring constant K, and damping coefficient D to a density error regression model RM-Δρ as a trained machine learning model, and estimates a density error Δρ of the fluid associated with the occurrence of adhesion. The diagnosis unit 15c also inputs at least one of frequency f, phase difference θ, signal amplitude A, mass M, spring constant K, and damping coefficient D to a flow rate error regression model RM-ΔQ as a trained machine learning model, and estimates a flow rate error ΔQ of the fluid associated with the occurrence of adhesion.

[0142] The diagnosing unit 15c inputs the calculated parameter change rate into the adhesion discrimination model DM-D as a trained machine learning model, and diagnoses the occurrence of adhesion on the vibration tube 13a as an abnormal state in the vibration tube 13a.

[0143] To explain a specific example of the adhesion diagnosis process, the diagnosis unit 15c inputs "frequency f001," "phase difference θ001," "signal amplitude A001," "mass M001," "spring constant K001," and "damping coefficient D001" to the adhesion discrimination model DM-D, obtains "adhesion present" output from the adhesion discrimination model DM-D in response to the input, and diagnoses that adhesion has occurred to the vibration tube 13a at "time T001."

[0144] Furthermore, the diagnosis unit 15c inputs "frequency f001," "phase difference θ001," "signal amplitude A001," "mass M001," "spring constant K001," and "damping coefficient D001" into the adhesion amount regression model RM-D, acquires "Xmm" output by the adhesion amount regression model RM-D in response to the input, and estimates that an adhesion amount of X [mm] has occurred on the vibration tube 13a at "time T001."

[0145] In addition, the diagnosis unit 15c inputs the “frequency f001,” “phase difference θ001,” “signal amplitude A001,” “mass M001,” “spring constant K001,” and “damping coefficient D001” into the density error regression model RM-Δρ, acquires “Δρ-D001” output by the density error regression model RM-Δρ in accordance with the input, and estimates the error in density ρ of “Δρ-D001” due to adhesion to the vibrating tube 13a at “time T001.”

[0146] In addition, the diagnosis unit 15c inputs the "frequency f001," "phase difference θ001," "signal amplitude A001," "mass M001," "spring constant K001," and "damping coefficient D001" into the flow error regression model RM-ΔQ, acquires "ΔQm-D001" output by the flow error regression model RM-ΔQ in response to the input, and estimates the error in the mass flow rate Qm of "ΔQm-D001" due to adhesion to the vibrating tube 13a at "time T001."

[0147] In addition, the diagnosis unit 15c inputs "mass M001" and the average change rate ΔM calculated by the calculation unit 15b into the corrosion discrimination model DM-R, obtains "adhesion present" output by the adhesion discrimination model DM-D in response to the input, and diagnoses that adhesion has occurred on the vibrating tube 13a at "time T001".

[0148] (Alarm notification processing) The diagnosis unit 15c executes an alarm notification process. For example, when the diagnosis unit 15c diagnoses the occurrence of an abnormal state in the vibration pipe 13a, the diagnosis unit 15c notifies an alarm to a terminal used by a user of the Coriolis flowmeter 10.

[0149] A specific example of the alarm notification process will be described. When diagnosing a corrosion abnormality, diagnosis unit 15c notifies operator terminal 20 of operator O who uses Coriolis flowmeter 10 identified by "Coriolis flowmeter CF001" of an alarm such as "A corrosion abnormality has been detected in Coriolis flowmeter CF001. Please replace the vibrating tube." When diagnosing an adhesion abnormality, diagnosis unit 15c notifies operator terminal 20 of operator O who uses Coriolis flowmeter 10 identified by "Coriolis flowmeter CF001" of an alarm such as "A adhesion abnormality has been detected in Coriolis flowmeter CF001. Please perform maintenance."

[0150] (2-3. Specific Examples of Each Process of the Coriolis Flowmeter 10) Specific examples of each process of the Coriolis flowmeter 10 will be described with reference to Fig. 8 to Fig. 10. Fig. 8 to Fig. 10 are diagrams showing specific examples of the diagnosis control process of the Coriolis flowmeter 10 according to the embodiment. Among the diagnosis control processes, specific example 1 using the discriminant model DM, specific example 2 utilizing a long-term trend, and specific example 3 using the regression model RM will be described below.

[0151] (2-3-1. Specific Example 1) 8, a specific example 1 using the discrimination model DM in the diagnosis control process of the Coriolis flowmeter 10 will be described. The discrimination model DM is a trained machine learning model that is trained using training data including at least one of the frequency f, the phase difference θ, the signal amplitude A, the mass M, the spring constant K, and the damping coefficient D, and is trained using the classification of corrosion or adhesion as teacher data.

[0152] In the example of Figure 8, the discrimination model DM receives at least one input of frequency f, phase difference θ, signal amplitude A, mass M, spring constant K, and damping coefficient D in the input layer, performs analysis of the received input data in the middle layer, and outputs diagnostic results such as "corrosion discrimination" and "adhesion discrimination" in the output layer.

[0153] A typical index for determining the presence or absence of corrosion is the spring constant K of the input layer. The Coriolis flowmeter 10 uses the spring constant K as an index to set a certain threshold value K A If the spring constant K falls below 0.05, it may be determined that corrosion has occurred. However, parameters such as the spring constant K may be affected by disturbances such as fluctuations in temperature and pressure of the fluid and the inclusion of air bubbles. Therefore, the Coriolis flowmeter 10 can make a determination by combining the spring constant K with other indices as a more reliable diagnostic control process.

[0154] In the example of FIG. 8, a neural network is shown as the discrimination model DM, but any number of layers and neurons can be set for the intermediate layers. In addition, a general nonlinear function such as a ramp function, a sigmoid function, or a hyperbolic tangent function is set as the activation function. From the viewpoint of the backpropagation method, it is preferable to provide a softmax function in the output layer when performing multi-value classification. In addition to a neural network, logistic regression, a support vector machine, a classification tree, change point detection, k-nearest neighbor method, and k-means method can be adopted as the discrimination model DM.

[0155] (2-3-2. Specific Example 2) Specific example 2 of the diagnosis control process of the Coriolis flowmeter 10, which is an applied diagnosis control process utilizing a long-term trend, will be described with reference to FIG.

[0156] As shown in FIG. 9, the Coriolis flowmeter 10 may be used as an index of corrosion or adhesion in combination with the long-term trend of the mass M. As shown in FIG. 9(1), even if no corrosion or adhesion occurs, the mass M fluctuates with a certain fluctuation range due to the influence of temperature, pressure, air bubbles, etc. The time scale of the horizontal axis of the graph in FIG. 9 is a relatively long time scale such as hours or days. Here, as shown in FIG. 9(2), when corrosion progresses, the piping mass of the vibrating pipe 13a decreases, so the mass gradually decreases in the long term. As shown in FIG. 9(3), the Coriolis flowmeter 10 may monitor the lower limit of the fluctuation range, and if the lower limit has a negative slope in the long-term trend, diagnose the occurrence of corrosion. The above monitor may be a method of recording the minimum value of the mass M in one day, for example.

[0157] As described above, Coriolis flowmeter 10 may diagnose the occurrence of corrosion or adhesion by classifying multiple parameters and long-term trend data of parameters in addition to spring constant K. Coriolis flowmeter 10 can also diagnose the occurrence of corrosion or adhesion by storing parameters in an initial state where no corrosion or adhesion occurs or in a state where adhesion has been cleaned off, and calculating the amount of change from the time of storage.

[0158] (2-3-3. Specific Example 3) 10, a specific example 3 using the regression model RM will be described among the diagnostic control processes of the Coriolis flowmeter 10. The regression model RM is a trained machine learning model trained using training data including at least one of the frequency f, phase difference θ, signal amplitude A, mass M, spring constant K, and damping coefficient D, and training data including the regression of the amount of corrosion or the amount of adhesion, or the regression of the density error or the flow rate error.

[0159] In the example of FIG. 10, the regression model RM receives at least one input of frequency f, phase difference θ, signal amplitude A, mass M, spring constant K, and damping coefficient D in the input layer, analyzes the received input data in the middle layer, and outputs diagnostic results such as “amount of corrosion / amount of adhesion,” “density correction / flow rate correction,” etc. in the output layer.

[0160] In the regression of the corrosion amount, a spring constant K is typically used. In the regression calculation, the corrosion amount is estimated from the correlation between the spring constant K and the corrosion amount. The corrosion amount may be notified to the operator O by a method such as notifying the operator O that the corrosion has progressed in mm or % with respect to the thickness of the vibration tube 13a. Normally, as the corrosion progresses, the spring constant K of the vibration tube 13a decreases and the inner diameter d also increases, so that an error occurs in the density ρ and the flow rate Q of the fluid measured by the Coriolis flowmeter 10. In the regression calculation of corrosion, the Coriolis flowmeter 10 may perform regression calculation of the error of the density ρ and the flow rate Q of the fluid from at least one of the parameters described in the input layer or the regression-calculated corrosion amount. The Coriolis flowmeter 10 not only notifies the operator O of the diagnosis result of the density error Δρ and the flow rate error ΔQ, but also uses it for correction. In addition, the Coriolis flowmeter 10 can set a threshold value of the corrosion amount or the density error Δρ and the flow rate error ΔQ by the operator O, so that an alarm may be notified when the result of the regression calculation exceeds the threshold value specified by the operator O. Here, since excessive corrosion of Coriolis flowmeter 10 may result in a risk of fluid leakage, a corrosion limit value is set for each product, and an alarm may be automatically issued to operator O when the limit value is exceeded.

[0161] On the other hand, in the regression of the adhesion amount, a damping coefficient D is typically used. In the regression calculation, the adhesion amount is estimated from the correlation between the damping coefficient D and the adhesion amount. However, unlike the case of corrosion, in the case of adhesion, physical property information of the adhesion (at least the density and Young's modulus of the adhesion) is required. The physical property information of the adhesion may be directly input by the operator O, or the adhesion can be estimated from the fluid type, so that information recorded in a data library may be used. In this way, when performing the regression calculation of the adhesion amount, the information of the adhesion must be given. The Coriolis flowmeter 10 may notify the operator O of the adhesion amount by a method such as notifying that the adhesion is progressing in mm or % with respect to the inner diameter d of the vibration tube 13a. In addition, the Coriolis flowmeter 10 may perform a regression calculation of the density error Δρ and the flow rate error ΔQ of the fluid, and not only notify the operator O of the result, but also perform correction. In addition, the notification of an alarm is the same as in the case of corrosion.

[0162] In the example of FIG. 10, a neural network is shown as the regression model RM, but any number of layers and neurons can be set for the intermediate layers. In addition, a general nonlinear function such as a ramp function, a sigmoid function, or a hyperbolic tangent function is set as the activation function. From the viewpoint of the backpropagation method, it is preferable to provide a softmax function in the output layer when performing multi-value classification. In addition to a neural network, multiple regression analysis, support vector regression, Gaussian process regression, regression tree, and autoregressive model can be adopted as the regression model RM.

[0163] (2-4. Configuration and Processing Examples of Operator Terminal 20) An example of the configuration and processing of operator terminal 20 will be described with reference to Fig. 2. Operator terminal 20 is a terminal used by operator O who performs maintenance, inspection, etc. of the plant and Coriolis flowmeter 10, and has an input / output unit 21, a transmission / reception unit 22, and a communication unit 23.

[0164] (2-4-1. Input / output section 21) The input / output unit 21 controls input of various information to the operator terminal 20. For example, the input / output unit 21 is realized by a mouse, a keyboard, a touch panel, or the like, and accepts input of setting information and the like to the operator terminal 20. In addition, the input / output unit 21 displays various information from the operator terminal 20. For example, the input / output unit 21 is realized by a display or the like, and displays setting information and the like stored in the operator terminal 20.

[0165] The input / output unit 21 displays an alarm notification screen to be viewed by the operator O. In addition, the input / output unit 21 displays an alarm notification time prediction screen to be viewed by the operator O.

[0166] (2-4-2. Transmitter / receiver 22) The transmitting / receiving unit 22 transmits various information. For example, the transmitting / receiving unit 22 transmits setting information and the like input by the operator O to the Coriolis flowmeter 10. The transmitting / receiving unit 22 receives various information. For example, the transmitting / receiving unit 22 receives an alarm notification, an alarm notification timing, and the like to be notified to the operator O from the Coriolis flowmeter 10.

[0167] (2-4-3. Communications Department 23) The communication unit 23 is responsible for data communication with other devices. For example, the communication unit 23 performs data communication with each communication device via a router, etc. The communication unit 23 can also perform data communication with a terminal (not shown).

[0168] (2-5. Configuration and Processing Examples of External Sensor 30) A configuration example and a processing example of the external sensor 30 will be described with reference to Fig. 2. The external sensor 30 is, for example, a pressure sensor, and is installed in a pipe P through which a fluid flows, and acquires upstream and downstream pressures P of the fluid. The external sensor 30 also transmits the acquired pressure P to the Coriolis flowmeter 10.

[0169] 3. Processing flow of the Coriolis flowmeter diagnostic system 100 11 to 14, a process flow of Coriolis flowmeter diagnostic system 100 according to an embodiment will be described. Below, the process flow of the entire Coriolis flowmeter diagnostic system 100 will be described, and then the acquisition control process, the calculation control process, and the diagnosis control process will be described.

[0170] (3-1. Overall processing of the Coriolis flowmeter diagnostic system 100) The overall process flow of the Coriolis flowmeter diagnostic system 100 according to the embodiment will be described with reference to Fig. 11. Fig. 11 is a flowchart showing an example of the overall process flow of the information processing system 100 according to the embodiment. Note that the processes of steps S101 to S103 below can also be executed in a different order. Also, some of the processes of steps S101 to S103 below may be omitted.

[0171] (3-1-1. Acquisition control process) First, Coriolis flowmeter 10 executes an acquisition control process (step S101). For example, Coriolis flowmeter 10 executes the processes of steps S201 to S205 described below to acquire the signal output by signal detector 13c as a measurement parameter.

[0172] (3-1-2. Calculation control process) Second, Coriolis flowmeter 10 executes a calculation control process (step S102). For example, Coriolis flowmeter 10 executes the processes of steps S301 to S310 described below to calculate vibration tube parameters indicating the state of vibration tube 13a.

[0173] (3-1-3. Diagnostic control processing) Third, Coriolis flowmeter 10 and operator terminal 20 execute a diagnosis control process (step S103). For example, Coriolis flowmeter 10 and operator terminal 20 execute the processes of steps S401 to S408 described below to execute a corrosion diagnosis or an adhesion diagnosis of Coriolis flowmeter 10 and notify operator O of an alarm.

[0174] (3-2. Measurement data management processing) The flow of the acquisition control process of the Coriolis flowmeter diagnostic system 100 according to the embodiment will be described with reference to Fig. 12. Fig. 12 is a flowchart showing an example of the flow of the acquisition control process of the Coriolis flowmeter diagnostic system 100 according to the embodiment. Note that the processes of steps S201 to S205 below may be executed in a different order. Also, some of the processes of steps S201 to S205 below may be omitted.

[0175] (3-2-1. Angular frequency acquisition process) First, the Coriolis flowmeter 10 executes an angular frequency acquisition process (step S201). For example, the Coriolis flowmeter 10 acquires the angular frequency ω from the output signal of the signal detector 13c.

[0176] (3-2-2. Voltage acquisition process) Second, the Coriolis flowmeter 10 executes a voltage acquisition process (step S202). For example, the Coriolis flowmeter 10 acquires a voltage ΔV from the output signal of the signal detector 13c.

[0177] (3-2-3. Frequency acquisition process) Third, Coriolis flowmeter 10 executes a frequency acquisition process (step S203) For example, Coriolis flowmeter 10 acquires frequency f from the output signal of signal detector 13c.

[0178] (3-2-4. Phase difference acquisition process) Fourth, the Coriolis flowmeter 10 executes a phase difference acquisition process (step S204). For example, the Coriolis flowmeter 10 acquires the phase difference θ from the output signal of the signal detector 13c.

[0179] (3-2-5. Signal amplitude acquisition process) Fifth, the Coriolis flowmeter 10 executes a signal amplitude acquisition process (step S205). For example, the Coriolis flowmeter 10 acquires the signal amplitude A from the output signal of the signal detector 13c.

[0180] (3-3. Calculation control process) A flow of calculation control processing in Coriolis flowmeter diagnostic system 100 according to the embodiment will be described with reference to Fig. 13. Fig. 13 is a flowchart showing an example of the flow of calculation control processing in Coriolis flowmeter diagnostic system 100 according to the embodiment. Note that the processing in steps S301 to S310 below can also be executed in a different order. Also, some of the processing in steps S301 to S310 below may be omitted.

[0181] (3-3-1. Fixed gain switching process 1) First, the Coriolis flowmeter 10 executes a fixed gain switching process 1 (step S301). For example, the Coriolis flowmeter 10 switches the oscillation feedback current gain G to G1.

[0182] (3-3-2. Square sum calculation process 1) Second, Coriolis flowmeter 10 executes sum-of-squares calculation process 1 (step S302). For example, Coriolis flowmeter 10 calculates the growth rate of the amplitude of the sine wave for each of periods i=1 to N.

[0183] (3-3-3. Calculation of average growth rate) Third, Coriolis flowmeter 10 executes a growth rate average value calculation process (step S303) For example, Coriolis flowmeter 10 calculates the average value of the growth rates calculated in the process of step S302.

[0184] (3-3-4. Fixed gain switching process 2) Fourth, the Coriolis flowmeter 10 executes a fixed gain switching process 2 (step S304). For example, the Coriolis flowmeter 10 switches the oscillation feedback current gain G to G2.

[0185] (3-3-5. Square sum calculation process 2) Fifth, Coriolis flowmeter 10 executes sum-of-squares calculation process 2 (step S305). For example, Coriolis flowmeter 10 calculates the attenuation rate of the amplitude of the sine wave for each of periods i=1 to N.

[0186] (3-3-6. Calculation of the average attenuation rate) Sixth, the Coriolis flowmeter 10 executes an attenuation rate average value calculation process (step S306) For example, the Coriolis flowmeter 10 calculates the average value of the attenuation rates calculated in the process of step S305.

[0187] (3-3-7. Switching process) Seventh, Coriolis flowmeter 10 executes a switch changeover process (step S307). For example, Coriolis flowmeter 10 changes switch 15a-6 from a side that selects fixed gain setting voltage E, which is the output value of voltage setting unit 15a-5, to a side that selects output signal ΔV of control amplifier 15a-4.

[0188] (3-3-8. Mass calculation process) Eighth, the Coriolis flowmeter 10 executes a mass calculation process (step S308). For example, the Coriolis flowmeter 10 calculates the total mass M of the vibrating tube 13a.

[0189] (3-3-9. Spring constant calculation process) Ninth, the Coriolis flowmeter 10 executes a spring constant calculation process (step S309). For example, the Coriolis flowmeter 10 calculates the spring constant K of the vibration tube 13a.

[0190] (3-3-10. Damping coefficient calculation process) Tenth, the Coriolis flowmeter 10 executes a damping coefficient calculation process (step S310). For example, the Coriolis flowmeter 10 calculates the damping coefficient D of the vibrating tube 13a.

[0191] (3-4. Diagnostic control processing) A flow of diagnostic control processing of Coriolis flowmeter diagnostic system 100 according to the embodiment will be described with reference to Fig. 13. Fig. 13 is a flowchart showing an example of the flow of diagnostic control processing of Coriolis flowmeter diagnostic system 100 according to the embodiment. Note that the processing of steps S401 to S408 below can also be executed in a different order. Also, some of the processing of steps S401 to S408 below may be omitted.

[0192] (3-4-1. Corrosion discrimination processing) First, the Coriolis flowmeter 10 executes a corrosion discrimination process (step S401). For example, the Coriolis flowmeter 10 diagnoses the occurrence of a corrosion abnormality by inputting parameters into a corrosion discrimination model DM-R. If a corrosion abnormality has occurred (step S402: Yes), the Coriolis flowmeter 10 proceeds to the process of step S403. On the other hand, if a corrosion abnormality has not occurred (step S402: No), the Coriolis flowmeter 10 proceeds to the process of step S406.

[0193] (3-4-2. Corrosion amount regression processing) Second, the Coriolis flowmeter 10 executes a corrosion amount regression process (step S403). For example, the Coriolis flowmeter 10 estimates the amount of corrosion by inputting parameters to a corrosion amount regression model RM-R.

[0194] (3-4-3. Corrosion alarm notification processing) Third, Coriolis flowmeter 10 executes a corrosion alarm notification process (step S404). For example, when Coriolis flowmeter 10 diagnoses the occurrence of corrosion, it notifies operator terminal 20 of a corrosion alarm and ends the diagnosis control process.

[0195] (3-4-4. Adhesion determination process) Fourth, the Coriolis flowmeter 10 executes an adhesion determination process (step S405). For example, the Coriolis flowmeter 10 diagnoses the occurrence of an adhesion abnormality by inputting parameters to an adhesion determination model DM-D. If an adhesion abnormality has occurred (step S406: Yes), the Coriolis flowmeter 10 proceeds to the process of step S407. On the other hand, if a corrosion abnormality has not occurred (step S406: No), the Coriolis flowmeter 10 ends the diagnosis control process.

[0196] (3-4-5. Corrosion regression processing) Fifth, the Coriolis flowmeter 10 executes adhesion amount regression processing (step S407). For example, the Coriolis flowmeter 10 estimates the adhesion amount by inputting parameters to the adhesion amount regression model RM-D.

[0197] (3-4-6. Adhesion alarm notification processing) Sixth, the Coriolis flowmeter 10 executes an adhesion alarm notification process (step S408). For example, when the Coriolis flowmeter 10 diagnoses the occurrence of adhesion, it notifies the operator terminal 20 of an adhesion alarm and ends the diagnosis control process.

[0198] 4. Effects of the embodiment Finally, effects of the embodiment will be described below: Effects 1 to 9 corresponding to the processing according to the embodiment will be described below.

[0199] (4-1. Effect 1) First, in the process according to the embodiment described above, the Coriolis flowmeter 10 acquires measurement parameters related to the measurement of the fluid, calculates vibration tube parameters indicating the state of the vibration tube 13a, and inputs at least one of the measurement parameters and the vibration tube parameters to a trained machine learning model that has been trained to output a diagnosis result for the vibration tube 13a in response to the input of the parameters, thereby diagnosing the occurrence of an abnormal state in the vibration tube 13a. Therefore, in this process, the occurrence of an abnormal state in the vibration tube 13a of the Coriolis flowmeter 10 can be effectively diagnosed.

[0200] (4-2. Effect 2) Secondly, in the process according to the embodiment described above, the Coriolis flowmeter 10 acquires the frequency f, phase difference θ, and signal amplitude A based on the vibration of the vibration tube 13a as measurement parameters, calculates the mass M, spring constant K, and damping coefficient D of the vibration tube 13a as vibration tube parameters, and inputs at least one of the frequency f, phase difference θ, signal amplitude A, mass M, spring constant K, and damping coefficient D to the corrosion discrimination model DM-R as a trained machine learning model to diagnose the occurrence of corrosion in the vibration tube 13a as an abnormal state in the vibration tube 13a. Therefore, in this process, the occurrence of a corrosion abnormality in the vibration tube 13a of the Coriolis flowmeter 10 can be effectively diagnosed.

[0201] (4-3. Effect 3) Thirdly, in the process according to the embodiment described above, the Coriolis flowmeter 10 inputs at least one of the frequency f, the phase difference θ, the signal amplitude A, the mass M, the spring constant K, and the damping coefficient D into the corrosion amount regression model RM-R as a trained machine learning model to estimate the corrosion amount of the vibrating tube 13a. Therefore, in this process, the corrosion abnormality in the vibrating tube 13a of the Coriolis flowmeter 10 can be quantitatively estimated, thereby effectively diagnosing the corrosion abnormality.

[0202] (4-4. Effect 4) Fourth, in the process according to the embodiment described above, Coriolis flowmeter 10 inputs at least one of frequency f, phase difference θ, signal amplitude A, mass M, spring constant K, and damping coefficient D to flow error regression model RM-ΔQ or density error regression model RM-Δρ as a trained machine learning model, and estimates at least one of flow error ΔQ and density error Δρ of the fluid associated with the occurrence of corrosion. Therefore, in this process, the occurrence of corrosion abnormality in vibration tube 13a of Coriolis flowmeter 10 can be effectively diagnosed, and measurement error can be corrected.

[0203] (4-5. Effect 5) Fifth, in the process according to the embodiment described above, the Coriolis flowmeter 10 acquires the frequency f, phase difference θ, and signal amplitude A based on the vibration of the vibrating tube 13a as measurement parameters, calculates the mass M, spring constant K, and damping coefficient D of the vibrating tube 13a as vibration tube parameters, and inputs at least one of the frequency f, phase difference θ, signal amplitude A, mass M, spring constant K, and damping coefficient D to the adhesion discrimination model DM-D as a trained machine learning model to diagnose the occurrence of adhesion on the vibrating tube 13a as an abnormal state in the vibrating tube 13a. Therefore, in this process, the occurrence of adhesion abnormality on the vibrating tube 13a of the Coriolis flowmeter 10 can be effectively diagnosed.

[0204] (4-6. Effect 6) Sixth, in the process according to the embodiment described above, the Coriolis flowmeter 10 inputs at least one of the frequency f, the phase difference θ, the signal amplitude A, the mass M, the spring constant K, and the damping coefficient D to the adhesion amount regression model RM-D as a trained machine learning model to estimate the adhesion amount of the vibrating tube 13a. Therefore, in this process, the adhesion abnormality in the vibrating tube 13a of the Coriolis flowmeter 10 can be effectively diagnosed by quantitatively estimating the adhesion abnormality.

[0205] (4-7. Effect 7) Seventh, in the process according to the embodiment described above, Coriolis flowmeter 10 inputs at least one of frequency f, phase difference θ, signal amplitude A, mass M, spring constant K, and damping coefficient D to flow error regression model RM-ΔQ or density error regression model RM-Δρ as a trained machine learning model to estimate at least one of flow rate error ΔQ and density error Δρ of the fluid associated with the occurrence of adhesion. Therefore, in this process, it is possible to effectively diagnose the occurrence of adhesion abnormality in vibration tube 13a of Coriolis flowmeter 10 and correct measurement errors.

[0206] (4-8. Effect 8) Eighth, in the process according to the embodiment described above, the Coriolis flowmeter 10 calculates the rate of change of a parameter for a predetermined period, inputs the rate of change into a trained machine learning model, and diagnoses the occurrence of an abnormal state in the vibrating tube 13a. Therefore, in this process, by utilizing long-term trend data, the occurrence of an abnormal state in the vibrating tube 13a of the Coriolis flowmeter 10 can be effectively diagnosed.

[0207] (4-9. Effect 9) Ninth, in the process according to the embodiment described above, when Coriolis flowmeter 10 diagnoses the occurrence of an abnormal state in vibrating tube 13a, it notifies an alarm to operator terminal 20 used by operator O, who is the user of Coriolis flowmeter 10. Therefore, in this process, it is possible to effectively diagnose the occurrence of an abnormal state in vibrating tube 13a of Coriolis flowmeter 10, and to have operator O perform maintenance.

[0208] [Modifications of the embodiment] Below, the configuration and processing of Coriolis flowmeter diagnostic system 100M according to the modified example of the embodiment, the configuration and processing of each device of Coriolis flowmeter diagnostic system 100M will be described in order, and finally, the effect of the modified example of the embodiment will be described. Note that the description of the configuration and processing common to the embodiment will be omitted.

[0209] [1. Configuration and Processing of Coriolis Flowmeter Diagnostic System 100M] The configuration and processing of Coriolis flowmeter diagnostic system 100M according to the embodiment will be described in detail with reference to Fig. 15. Fig. 15 is a diagram showing a configuration example and processing example of Coriolis flowmeter diagnostic system 100M according to a modified embodiment. Below, a configuration example of the entire Coriolis flowmeter diagnostic system 100M and a processing example of the entire Coriolis flowmeter diagnostic system 100M will be described.

[0210] (1-1. Example of the overall configuration of the Coriolis flowmeter diagnostic system 100M) An example of the overall configuration of Coriolis flowmeter diagnostic system 100M will be described with reference to Fig. 15. Coriolis flowmeter diagnostic system 100M has Coriolis flowmeter 10M, operator terminal 20, and server device 40. Coriolis flowmeter 10M, operator terminal 20, and server device 40 are connected to each other via a predetermined communication network (not shown) so as to be able to communicate with each other by wire or wirelessly. The predetermined communication network may be any of various communication networks such as the Internet or a dedicated line.

[0211] (1-1-1. Coriolis flowmeter 10M) The Coriolis flowmeter 10M is a plant device that is installed at a plant site and measures the flow rate (mass flow rate, volumetric flow rate) and density of a fluid flowing through a piping P of the plant. Note that the Coriolis flowmeter diagnostic system 100M shown in FIG. 15 may include a plurality of Coriolis flowmeters 10M.

[0212] (1-1-2. Operator terminal 20) The operator terminal 20 is an information processing device used by an operator O. Note that the Coriolis flowmeter diagnostic system 100M shown in Fig. 15 may include a plurality of operator terminals 20. Note that the operator terminal 20 is common to the Coriolis flowmeter diagnostic system 100 described above, and therefore a description thereof will be omitted.

[0213] (1-1-3. Server device 40) Server device 40 is an information processing device that diagnoses the occurrence of an abnormal state in Coriolis flowmeter 10M. Note that Coriolis flowmeter diagnostic system 100M shown in Fig. 15 may include server device 40. Also, server device 40 may be realized by a cloud environment, an on-premise environment, an edge environment, or the like.

[0214] (1-2. Example of overall processing for Coriolis flowmeter diagnostic system 100M) An example of the overall processing of Coriolis flowmeter diagnostic system 100M will be described with reference to Fig. 15. Note that the processing of steps S11 to S16 below may be executed in a different order. Also, some of the processing of steps S11 to S16 below may be omitted.

[0215] (1-2-1. Measurement parameter acquisition process) First, Coriolis flowmeter 10M acquires measurement parameters (step S11). For example, Coriolis flowmeter 10M acquires angular frequency ω and voltage ΔV corresponding to the vibration of vibrating tube 13a (not shown) of Coriolis flowmeter 10 from signal detector 13c (not shown) of Coriolis flowmeter 10M. Coriolis flowmeter 10M also acquires frequency f, phase difference θ, and signal amplitude A corresponding to the vibration of vibrating tube 13a of Coriolis flowmeter 10M from signal detector 13c of Coriolis flowmeter 10M.

[0216] (1-2-2. Vibration tube parameter calculation process) Second, the Coriolis flowmeter 10M calculates the vibrating tube parameters (step S12). For example, the Coriolis flowmeter 10M calculates the growth rate and the damping rate of the vibrating tube 13a, and calculates the mass M, the spring constant K, and the damping coefficient D of the vibrating tube 13a.

[0217] (1-2-3. Parameter transmission process) Third, Coriolis flowmeter 10M transmits parameters to server device 40 (step S13). For example, Coriolis flowmeter 10M transmits frequency f, phase difference θ, signal amplitude A, mass M, spring constant K, damping coefficient D, and the like to server device 40 as parameters.

[0218] (1-2-4. Corrosion diagnosis processing) Fourth, the server device 40 diagnoses the corrosion of the vibration pipe 13a (step S14). For example, the server device 40 inputs at least one of the frequency f, the phase difference θ, the signal amplitude A, the mass M, the spring constant K, and the damping coefficient D to a corrosion discrimination model DM-R, which is a trained machine learning model, to discriminate the occurrence of corrosion of the vibration pipe 13a.

[0219] In addition, the server device 40 inputs at least one of the frequency f, the phase difference θ, the signal amplitude A, the mass M, the spring constant K, and the damping coefficient D into the corrosion amount regression model RM-R, which is a trained machine learning model, to estimate the corrosion amount of the vibration tube 13a.

[0220] In addition, the server device 40 inputs at least one of the frequency f, phase difference θ, signal amplitude A, mass M, spring constant K, and damping coefficient D into a density error regression model RM-Δρ, which is a trained machine learning model, to estimate the density error Δρ associated with the corrosion of the vibrating tube 13a.

[0221] In addition, the server device 40 inputs at least one of the frequency f, phase difference θ, signal amplitude A, mass M, spring constant K, and damping coefficient D into a flow error regression model RM-ΔQ, which is a trained machine learning model, to estimate the flow error ΔQ associated with corrosion of the vibrating tube 13a.

[0222] (1-2-5. Adhesion diagnosis processing) Fifth, the server device 40 diagnoses adhesion of the vibration pipe 13a (step S15). For example, the server device 40 inputs at least one of the frequency f, the phase difference θ, the signal amplitude A, the mass M, the spring constant K, and the damping coefficient D to an adhesion discrimination model DM-D, which is a trained machine learning model, to determine the occurrence of adhesion of the vibration pipe 13a.

[0223] In addition, the server device 40 inputs at least one of the frequency f, phase difference θ, signal amplitude A, mass M, spring constant K, and damping coefficient D into the adhesion amount regression model RM-D, which is a trained machine learning model, to estimate the adhesion amount on the vibrating tube 13a.

[0224] In addition, the server device 40 inputs at least one of the frequency f, phase difference θ, signal amplitude A, mass M, spring constant K, and damping coefficient D into a density error regression model RM-Δρ, which is a trained machine learning model, to estimate the density error Δρ associated with the adhesion of the vibrating tube 13a.

[0225] In addition, the server device 40 inputs at least one of the frequency f, phase difference θ, signal amplitude A, mass M, spring constant K, and damping coefficient D into a flow error regression model RM-ΔQ, which is a trained machine learning model, to estimate the flow error ΔQ due to the adhesion of the vibrating tube 13a.

[0226] (1-2-6. Alarm notification processing) Sixth, the server device 40 notifies the operator terminal 20 of an alarm (step S16). For example, the server device 40 notifies the operator terminal 20 of an alarm indicating the occurrence of corrosion or adhesion of the vibrating tube 13a, and prompts the operator O to perform maintenance of the Coriolis flowmeter 10.

[0227] Furthermore, the server device 40 notifies the operator terminal 20 of an alarm including the amount of corrosion or the amount of adhesion of the vibration pipe 13a.

[0228] Furthermore, the server device 40 notifies the operator terminal 20 of an alarm including the density error Δρ or the flow rate error ΔQ associated with the corrosion or adhesion of the vibration pipe 13a.

[0229] [2. Configuration and processing of each device of the Coriolis flowmeter diagnostic system 100M] The configuration and processing of each device of Coriolis flowmeter diagnostic system 100M shown in FIG. 15 will be described with reference to FIG. 16. FIG. 16 is a block diagram showing an example of the configuration of each device of Coriolis flowmeter diagnostic system 100M according to a modified example of the embodiment. In the following, an example of the configuration of Coriolis flowmeter diagnostic system 100M as a whole according to a modified example of the embodiment will be described, and then an example of the configuration and processing of Coriolis flowmeter 10M according to the modified example of the embodiment, and an example of the configuration and processing of server device 40 will be described in detail. Note that operator terminal 20 and external sensor 30 are common to Coriolis flowmeter diagnostic system 100 described above, and therefore description thereof will be omitted.

[0230] (2-1. Example of the overall configuration of the Coriolis flowmeter diagnostic system 100M) An example of the overall configuration of Coriolis flowmeter diagnostic system 100M shown in FIG. 15 will be described with reference to FIG. 16. As shown in FIG. 16, Coriolis flowmeter diagnostic system 100M has Coriolis flowmeter 10M, operator terminal 20, external sensor 30, and server device 40. Coriolis flowmeter 10M is communicatively connected to operator terminal 20, external sensor 30, and server device 40 via a communication network N realized by the Internet, a dedicated line, or the like. Server device 40 may be installed in a cloud environment, an on-premise environment, an edge environment, or the like. Furthermore, external sensor 30 is installed in a plant.

[0231] (2-2. Example of configuration and processing of Coriolis flowmeter 10M) A configuration example and a processing example of the Coriolis flowmeter 10M will be described with reference to Fig. 16. The Coriolis flowmeter 10M is a measuring device that measures the flow rate and density of a fluid flowing through the vibrating tube 13a based on the vibration of the vibrating tube 13a, and has an input unit 11, an output unit 12, a detection unit 13, and a control unit 15M. Note that the input unit 11, the output unit 12, and the detection unit 13 are common to the above-mentioned Coriolis flowmeter 10, and therefore description thereof will be omitted.

[0232] (2-2-1. Control unit 15M) Control unit 15M controls the entire Coriolis flowmeter 10M. Control unit 15 has an acquisition unit 15a and a calculation unit 15b. Note that acquisition unit 15a and calculation unit 15b are common to Coriolis flowmeter 10 described above, and therefore description thereof will be omitted.

[0233] (2-3. Configuration and Processing Examples of Server Device 40) A configuration example and a processing example of the server device 40 will be described with reference to Fig. 16. The server device 40 has an input unit 41, an output unit 42, a communication unit 43, a storage unit 44, and a control unit 45. Note that the storage unit 44 is common to the storage unit 14 of the Coriolis flowmeter 10 described above, and therefore a description thereof will be omitted.

[0234] (2-3-1. Input section 41) The input unit 41 controls input of various information to the server device 40. For example, the input unit 41 is realized by a mouse, a keyboard, etc., and accepts input of various information to the server device 40.

[0235] (2-3-2. Output section 42) The output unit 42 controls the output of various information from the server device 40. For example, the output unit 42 is realized by a display or the like, and displays various information stored in the server device 40.

[0236] (2-3-3. Communications Department 43) The communication unit 43 manages data communication with other devices. For example, the communication unit 43 performs data communication with each communication device via a router, etc. The communication unit 43 can also perform data communication with an operator's terminal (not shown).

[0237] (2-3-4. Control unit 45) The control unit 45 controls the entire server device 40. The control unit 45 has a receiving unit 45a and a diagnosing unit 45b. Note that the diagnosing unit 45b is common to the diagnosing unit 15c of the Coriolis flowmeter 10 described above, and therefore a description thereof will be omitted.

[0238] (2-2-4-1. Receiving unit 45a) The receiver 45a receives various information. The receiver 45a stores the received various information in the storage unit 44. For example, the receiver 45a receives measurement parameters and vibration tube parameters. At this time, the receiver 45a receives the frequency f, phase difference θ, and signal amplitude A based on the vibration of the vibration tube 13a from the Coriolis flowmeter 10M as the measurement parameters. The receiver 45a stores the received measurement parameters in the acquired value storage unit 44a. The receiver 45a also receives the mass M, spring constant K, and damping coefficient D of the vibration tube 13a as the vibration tube parameters. The receiver 45a stores the received vibration tube parameters in the calculated value storage unit 44b.

[0239] 3. Effects of the Modifications of the Embodiments Finally, the effect of the modified embodiment will be described. In the process according to the modified embodiment described above, the Coriolis flowmeter 10M acquires measurement parameters related to the measurement of the fluid of the Coriolis flowmeter 10M, calculates vibration tube parameters indicating the state of the vibration tube 13a, and the server device 40 receives the measurement parameters and the vibration tube parameters, and inputs at least one of the measurement parameters and the vibration tube parameters to a trained machine learning model trained to output a diagnosis result of the vibration tube 13a according to the input of the parameters, thereby diagnosing the occurrence of an abnormal state in the vibration tube 13a. Therefore, in this process, the occurrence of an abnormal state in the vibration tube 13a of the Coriolis flowmeter 10 can be effectively diagnosed by the server device 40 having a large resource executing the diagnosis control process.

[0240] 〔system〕 The information including the processing procedures, control procedures, specific names, various data and parameters shown in the above documents and drawings can be changed arbitrarily unless otherwise specified.

[0241] In addition, each component of each device shown in the figure is a functional concept, and does not necessarily have to be physically configured as shown in the figure. In other words, the specific form of distribution and integration of each device is not limited to that shown in the figure. In other words, all or part of them can be functionally or physically distributed and integrated in any unit according to various loads, usage conditions, etc.

[0242] Furthermore, each processing function performed by each device may be realized, in whole or in part, by a CPU and a program analyzed and executed by the CPU, or may be realized as hardware using wired logic.

[0243] [Hardware] Next, a hardware configuration example of a Coriolis flowmeter 10 that is an information management device will be described. Note that other devices may have a similar hardware configuration. FIG. 17 is a diagram showing a hardware configuration example according to an embodiment. As shown in FIG. 17, the Coriolis flowmeter 10 has a communication device 10a, an EEPROM (Electrically Erasable Programmable Read-Only Memory) 10b, a memory 10c, and a processor 10d. In addition, each unit shown in FIG. 17 is connected to each other by a bus or the like.

[0244] The communication device 10a is a network interface card or the like, and communicates with other servers. The EEPROM 10b stores programs and databases that operate the functions shown in FIG.

[0245] The processor 10d reads out a program that executes the same processes as the respective processing units shown in Fig. 2 from the EEPROM 10b etc. and loads it in the memory 10c, thereby operating a process that executes the respective functions described in Fig. 2 etc. For example, this process executes the same functions as the respective processing units of the Coriolis flowmeter 10. Specifically, the processor 10d reads out a program having the same functions as the acquisition unit 15a, the calculation unit 15b, the diagnosis unit 15c etc. from the EEPROM 10b etc. Then, the processor 10d executes a process that executes the same processes as the acquisition unit 15a, the calculation unit 15b, the diagnosis unit 15c etc.

[0246] In this way, Coriolis flowmeter 10 operates as a device that executes various processing methods by reading and executing a program. Coriolis flowmeter 10 can also realize functions similar to those of the above-mentioned embodiment by reading the program from a recording medium using a medium reading device and executing the read program. Note that the program in these other embodiments is not limited to being executed by Coriolis flowmeter 10. For example, the present invention can be similarly applied to cases where another computer or server executes a program, or where these cooperate to execute a program.

[0247] This program can be distributed via a network such as the Internet. This program can also be recorded on a computer-readable recording medium such as a hard disk, a flexible disk (FD), a CD-ROM, a magneto-optical disk (MO), or a digital versatile disk (DVD), and can be executed by being read out from the recording medium by a computer.

[0248] 〔others〕 Some examples of combinations of the disclosed technical features are set out below.

[0249] (1) A Coriolis flowmeter that measures a flow rate and density of a fluid flowing through a vibrating tube based on the vibration of the vibrating tube, comprising: an acquisition unit that acquires measurement parameters related to the measurement of the fluid by the Coriolis flowmeter; a calculation unit that calculates vibrating tube parameters that indicate a state of the vibrating tube; and a diagnosis unit that inputs at least one of the measurement parameters and the vibrating tube parameters to a trained machine learning model that has been trained to output a diagnosis result of the vibrating tube in response to the input of parameters, and diagnoses the occurrence of an abnormal condition in the vibrating tube.

[0250] (2) The Coriolis flowmeter described in (1), wherein the acquisition unit acquires a frequency, a phase difference, and a signal amplitude based on the vibration of the vibrating tube as the measurement parameters, the calculation unit calculates a mass, a spring constant, and a damping coefficient of the vibrating tube as the vibrating tube parameters, and the diagnosis unit inputs at least one of the frequency, the phase difference, the signal amplitude, the mass, the spring constant, and the damping coefficient to the trained machine learning model to diagnose the occurrence of corrosion of the vibrating tube as an abnormal condition in the vibrating tube.

[0251] (3) The Coriolis flowmeter described in (2), wherein the diagnosis unit inputs at least one of the frequency, the phase difference, the signal amplitude, the mass, the spring constant, and the damping coefficient to the trained machine learning model to estimate the amount of corrosion of the vibrating tube.

[0252] (4) A Coriolis flowmeter as described in (2) or (3), wherein the diagnosis unit inputs at least one of the frequency, the phase difference, the signal amplitude, the mass, the spring constant, and the damping coefficient to the trained machine learning model to estimate at least one of a flow rate error and a density error of the fluid associated with the occurrence of the corrosion.

[0253] (5) The Coriolis flowmeter according to (1) or (2), wherein the acquisition unit acquires a frequency, a phase difference, and a signal amplitude as the measurement parameters, the calculation unit calculates a mass, a spring constant, and a damping coefficient of the vibration tube as the vibration tube parameters, and the diagnosis unit inputs at least one of the frequency, the phase difference, the signal amplitude, the mass, the spring constant, and the damping coefficient to the trained machine learning model to diagnose the occurrence of adhesion on the vibration tube as an abnormal condition in the vibration tube.

[0254] (6) The Coriolis flowmeter described in (5), wherein the diagnosis unit inputs at least one of the frequency, the phase difference, the signal amplitude, the mass, the spring constant, and the damping coefficient to the trained machine learning model to estimate an amount of adhesion on the vibrating tube.

[0255] (7) The Coriolis flowmeter according to (5) or (6), wherein the diagnostic unit inputs at least one of the frequency, the phase difference, the signal amplitude, the mass, the spring constant, and the damping coefficient into the trained machine learning model to estimate at least one of a flow rate error and a density error of the fluid associated with the occurrence of the adhesion.

[0256] (8) A Coriolis flowmeter described in any one of (1) to (7), wherein the calculation unit calculates a rate of change of a parameter over a predetermined period of time, and the diagnosis unit inputs the rate of change into the trained machine learning model to diagnose the occurrence of an abnormal condition in the vibrating tube.

[0257] (9) The Coriolis flowmeter according to any one of (1) to (8), wherein when the diagnostic unit diagnoses the occurrence of an abnormal state in the vibrating tube, the diagnostic unit notifies an alarm to a terminal used by a user of the Coriolis flowmeter.

[0258] (10) A diagnostic method in which a Coriolis flowmeter that measures the flow rate and density of a fluid flowing through a vibrating tube based on the vibration of the vibrating tube acquires measurement parameters related to the measurement of the fluid by the Coriolis flowmeter, calculates vibrating tube parameters that indicate the state of the vibrating tube, and inputs at least one of the measurement parameters and the vibrating tube parameters into a trained machine learning model that has been trained to output a diagnosis result of the vibrating tube in response to the input of the parameters, thereby diagnosing the occurrence of an abnormal condition in the vibrating tube.

[0259] (11) A diagnostic program that causes a Coriolis flowmeter that measures the flow rate and density of a fluid flowing through a vibrating tube based on the vibration of the vibrating tube to execute a process of acquiring measurement parameters related to the measurement of the fluid by the Coriolis flowmeter, calculating vibrating tube parameters that indicate the state of the vibrating tube, and inputting at least one of the measurement parameters and the vibrating tube parameters into a trained machine learning model that has been trained to output a diagnosis result of the vibrating tube in response to the input of the parameters, thereby diagnosing the occurrence of an abnormal condition in the vibrating tube.

[0260] (12) A diagnostic system including a Coriolis flowmeter that measures a flow rate and density of a fluid flowing through a vibrating tube based on the vibration of the vibrating tube, and a diagnostic device, wherein the Coriolis flowmeter includes an acquisition unit that acquires measurement parameters related to the measurement of the fluid by the Coriolis flowmeter, and a calculation unit that calculates vibrating tube parameters that indicate a state of the vibrating tube, and the diagnostic device includes a receiving unit that receives the measurement parameters and the vibrating tube parameters, and a diagnostic unit that inputs at least one of the measurement parameters and the vibrating tube parameters to a trained machine learning model that is trained to output a diagnosis result of the vibrating tube in response to the input of the parameters, to diagnose the occurrence of an abnormal condition in the vibrating tube. [Explanation of symbols]

[0261] 10, 10M Coriolis flowmeter 11, 41 Input section 12, 42 Output section 13 Detection section 13a Vibration tube 13b Shaker 13c Signal Detector 14, 44 Storage section 14a, 44a Acquired value storage section 14b, 44b Calculation value storage section 14c, 44c Diagnosis result storage section 14d, 44d Discriminant model memory section 14e, 44e Regression model memory section 15 Control section 15a Acquisition part 15b Calculation part 15c, 45b Diagnostics Department 20 Operator terminal 21 Input / output section 22 Transmitter / receiver 23 Communications Department 30 External Sensor 40 Server device 43 Communications Department 45 Control section 45a Receiving section 100, 100M Coriolis Flowmeter Diagnostic System

Claims

1. A Coriolis flowmeter for measuring a flow rate and a density of a fluid flowing through a vibrating tube based on the vibration of the vibrating tube, comprising: an acquisition unit that acquires measurement parameters related to the measurement of a fluid by the Coriolis flowmeter; A calculation unit that calculates a vibration tube parameter indicating a state of the vibration tube; a diagnosis unit that inputs at least one of the measurement parameters and the vibration pipe parameters into a trained machine learning model that has been trained to output a diagnosis result for the vibration pipe in response to an input of a parameter, and diagnoses the occurrence of an abnormal state in the vibration pipe; A Coriolis flowmeter comprising:

2. The acquisition unit is Acquiring a frequency, a phase difference, and a signal amplitude based on the vibration of the vibrating tube as the measurement parameters; The calculation unit is Calculating the mass, spring constant, and damping coefficient of the vibration tube as the vibration tube parameters; The diagnosis unit includes: inputting at least one of the frequency, the phase difference, the signal amplitude, the mass, the spring constant, and the damping coefficient into the trained machine learning model, and diagnosing the occurrence of corrosion of the vibration tube as an abnormal state in the vibration tube; 2. The Coriolis flowmeter of claim 1.

3. The diagnosis unit includes: At least one of the frequency, the phase difference, the signal amplitude, the mass, the spring constant, and the damping coefficient is input to the trained machine learning model to estimate the amount of corrosion of the vibration tube.

3. The Coriolis flowmeter of claim 2.

4. The diagnosis unit includes: inputting at least one of the frequency, the phase difference, the signal amplitude, the mass, the spring constant, and the damping coefficient into the trained machine learning model to estimate at least one of a flow rate error and a density error of the fluid due to the occurrence of the corrosion; 4. The Coriolis flowmeter of claim 3.

5. The acquisition unit is Acquire frequency, phase difference, and signal amplitude as the measurement parameters; The calculation unit is Calculating the mass, spring constant, and damping coefficient of the vibration tube as the vibration tube parameters; The diagnosis unit includes: inputting at least one of the frequency, the phase difference, the signal amplitude, the mass, the spring constant, and the damping coefficient into the trained machine learning model, and diagnosing the occurrence of adhesion on the vibration tube as an abnormal state in the vibration tube; 2. The Coriolis flowmeter of claim 1.

6. The diagnosis unit includes: At least one of the frequency, the phase difference, the signal amplitude, the mass, the spring constant, and the damping coefficient is input to the trained machine learning model to estimate the amount of adhesion on the vibration tube.

6. The Coriolis flowmeter of claim 5.

7. The diagnosis unit includes: inputting at least one of the frequency, the phase difference, the signal amplitude, the mass, the spring constant, and the damping coefficient into the trained machine learning model to estimate at least one of a flow rate error and a density error of the fluid caused by the occurrence of the adhesion; 7. The Coriolis flowmeter of claim 6.

8. The calculation unit is Calculating the rate of change of the parameter over a predetermined period of time; The diagnosis unit includes: The rate of change is input to the trained machine learning model to diagnose the occurrence of an abnormal state in the vibrating pipe. The Coriolis flowmeter according to claim 1 .

9. The diagnosis unit includes: When an abnormal state is diagnosed in the vibrating tube, an alarm is sent to a terminal used by a user of the Coriolis flowmeter. The Coriolis flowmeter according to claim 1 .

10. A Coriolis flowmeter for measuring a flow rate and a density of a fluid flowing through a vibrating tube based on the vibration of the vibrating tube, Obtaining measurement parameters relating to a measurement of the Coriolis flowmeter; Calculating a vibration tube parameter indicating a state of the vibration tube; inputting at least one of the measurement parameters and the vibration pipe parameters into a trained machine learning model trained to output a diagnosis result for the vibration pipe in response to the input of parameters, thereby diagnosing the occurrence of an abnormal state in the vibration pipe; A diagnostic method to perform the action.

11. A Coriolis flowmeter for measuring a flow rate and a density of a fluid flowing through a vibrating tube based on the vibration of the vibrating tube, Obtaining a measurement parameter relating to a measurement of a fluid by the Coriolis flowmeter; Calculating a vibration tube parameter indicating a state of the vibration tube; inputting at least one of the measurement parameters and the vibration pipe parameters into a trained machine learning model trained to output a diagnosis result for the vibration pipe in response to the input of parameters, thereby diagnosing the occurrence of an abnormal state in the vibration pipe; A diagnostic program that performs the process.

12. A diagnostic system including a Coriolis flowmeter that measures a flow rate and a density of a fluid flowing through a vibrating tube based on vibration of the vibrating tube, and a diagnostic device, The Coriolis flowmeter includes: an acquisition unit that acquires measurement parameters related to the measurement of a fluid by the Coriolis flowmeter; A calculation unit is provided for calculating a vibration tube parameter indicating a state of the vibration tube, The diagnostic device comprises: a receiving unit for receiving the measurement parameters and the vibration tube parameters; and a diagnosis unit that inputs at least one of the measurement parameters and the vibration pipe parameters to a trained machine learning model that has been trained to output a diagnosis result for the vibration pipe in response to an input of a parameter, and diagnoses the occurrence of an abnormal state in the vibration pipe. Diagnostic system.

Images