Evaluation method, program, evaluation device, and liquid material discharge device

Abstract

To provide an evaluation method, a program, an evaluation device, and a liquid material discharge device that can easily acquire more appropriate information on a liquid material discharged from the liquid material discharge device.SOLUTION: An evaluation method relates to a liquid material discharged from a liquid material discharge device that comprises: a driving unit (23); and a recording element (21) that has a piezoelectric element (213), outputs a driving signal from the driving unit (23) to displace the piezoelectric element (213) thereby discharging the liquid material, and applies the liquid material onto a medium to perform printing. At least one of the characteristics of the liquid material, discharge characteristics, and printing quality is estimated on the basis of information on the waveform of an electric signal detected on an input side of the driving signal to the piezoelectric element (213).SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present invention relates to an evaluation method, a program, an evaluation device, and a liquid material ejection device. [Background technology]

[0002] A liquid ejection device that ejects a liquid in the form of fine droplets to form an image, a coating, a structure, etc. on a medium can be used with a wide variety of liquids and media. For example, in an inkjet printer, which is a typical example of a liquid ejection device, ink, which is a liquid, must be ejected appropriately to obtain a good printed product.

[0003] Conventionally, there is a technique for controlling the discharge operation of a liquid material according to a parameter that is considered to be particularly important for the proper discharge of the liquid material. Viscosity is known as such a parameter. Viscosity can be determined electrically (for example, Patent Documents 1 and 2). [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Laid-Open No. 2005-61871 [Patent Document 2] Japanese Patent Application Laid-Open No. 2009-174930 Summary of the Invention [Problem to be solved by the invention]

[0005] However, there are many parameters that can affect the ejection of liquids. In addition to the viscosity mentioned above, other parameters include elasticity, density, specific gravity, and surface tension. Therefore, adjusting the liquid by focusing only on a few representative parameters, as in the past, does not necessarily result in proper ejection, and as a result, prints of sufficiently high quality are not obtained. Meanwhile, manually analyzing and quantitatively exploring the effects of numerous parameters to find the appropriate conditions can be a huge effort. Furthermore, unless the parameters themselves reflect the actual ink, device, and equipment used during printing, it is difficult to obtain accurate results.

[0006] An object of the present invention is to provide an evaluation method, a program, an evaluation device, and a liquid-material ejection device that can easily acquire more appropriate information about a liquid material ejected from the liquid-material ejection device. [Means for solving the problem]

[0007] In order to achieve the above object, one aspect of the present invention is 1. An evaluation method for a liquid material discharged by a liquid material discharge device, the evaluation method comprising: a drive unit; and a recording element having an electromechanical conversion element, the recording element displacing the liquid material by outputting a drive signal from the drive unit to displace the electromechanical conversion element, and applying the liquid material onto a medium to perform printing, the method comprising: At least one of the characteristics, ejection characteristics, and print quality of the liquid material is estimated based on waveform information of an electric signal detected at the input side of the drive signal to the electromechanical conversion element.

[0008] Another aspect of the present invention is an acquisition means for acquiring waveform information of an electric signal detected on the input side of the drive signal to the electromechanical conversion element from a liquid discharge device including a drive unit and a recording element having an electromechanical conversion element, the recording element outputting a drive signal from the drive unit to displace the electromechanical conversion element, thereby discharging a liquid material and applying the liquid material onto a medium to perform printing; an estimation means for estimating at least one of the characteristics, ejection characteristics, and print quality of the liquid material based on the waveform information; It is a program that functions as a

[0009] Another aspect of the present invention is an information acquiring unit that acquires waveform information of an electrical signal detected on the input side of the drive signal to the electromechanical conversion element from a liquid discharge device that includes a drive unit and a recording element that has an electromechanical conversion element and outputs a drive signal from the drive unit to displace the electromechanical conversion element to discharge a liquid material and apply the liquid material onto a medium to perform printing; an estimation unit that estimates at least one of the characteristics, ejection characteristics, and print quality of the liquid material based on the waveform information; The evaluation device is provided with:

[0010] Another aspect of the present invention is a drive unit that outputs a first drive signal; a liquid supply unit that supplies a known liquid; a recording element having an electromechanical conversion element, displacing the electromechanical conversion element in response to the first drive signal output from the drive unit to eject the known liquid material and apply the known liquid material onto a medium to perform printing; an information acquiring unit that acquires first waveform information of a first electrical signal detected at an input side of the drive signal to the electromechanical conversion element; a database generation unit that associates the first waveform information with at least one of the characteristics, ejection characteristics, and print quality of the known liquid material, and generates a database for making an estimation related to a liquid material to be evaluated that is different from the known liquid material; The liquid material ejection device includes: [Effects of the Invention]

[0011] According to the present invention, it is possible to easily obtain more appropriate information about the liquid material to be discharged from the liquid material discharge device. [Brief explanation of the drawings]

[0012] [Figure 1]FIG. 2 is a block diagram illustrating a functional configuration of the inkjet printing apparatus. [Figure 2] FIG. 2 is a diagram illustrating a recording operation unit. [Figure 3] 10 is a flowchart showing a control procedure for a database generation process. [Figure 4] 10 is a flowchart showing a control procedure for ink physical property estimation processing. [Figure 5] 10 is a flowchart showing a control procedure for a discharge characteristic estimation process. [Figure 6] 10 is a flowchart showing a control procedure for print quality estimation processing. [Figure 7] FIG. 10 is a diagram illustrating a recording operation unit according to a second embodiment. [Figure 8] FIG. 10 is a diagram illustrating a recording operation unit according to a third embodiment. [Figure 9] FIG. 10 is a diagram illustrating a recording operation unit according to a fourth embodiment. [Figure 10] FIG. 13 is a diagram illustrating a recording operation unit according to a fifth embodiment. [Figure 11] 13 is a flowchart showing a control procedure for database generation processing in an inkjet recording apparatus according to a fifth embodiment. [Figure 12] FIG. 13 is a diagram illustrating a recording operation unit according to a sixth embodiment. [Figure 13] 10 is a flowchart illustrating a control procedure for a print quality adjustment process according to an embodiment. [Figure 14] 10 is a flowchart illustrating a control procedure for a print quality adjustment process according to an embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0013] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing the functional configuration of an inkjet recording apparatus 1 including a liquid material ejection device according to this embodiment.

[0014] The inkjet recording device 1 includes a control unit 11, a memory unit 12, a communication unit 13, a display unit 14, an operation reception unit 15, a measurement unit 16, a recording operation unit 20, a discharge state observation unit 31, and the like.

[0015] The control unit 11 controls the overall operation of the inkjet recording apparatus 1. The control unit 11 includes a hardware processor. The hardware processor may be a general-purpose CPU (Central Processing Unit) or a dedicated processor. The hardware processor may be a single processor or multiple processors operating in parallel or independently.

[0016] The storage unit 12 has a RAM and a non-volatile memory. The RAM provides a working memory space for the hardware processor and stores primary data. The non-volatile memory stores a program 121 and a database 122. The non-volatile memory may include, for example, an HDD (Hard Disk Drive) and / or a flash memory. At least the control unit 11 and the storage unit 12 are included in the computer and evaluation device of this embodiment.

[0017] The program 121 includes a program for generating a database relating to the correspondence between the electrical signals related to the ejection operation of ink, which is the liquid material of this embodiment, and the ink characteristics (ink physical properties, ejection characteristics) and / or print quality. The program 121 also includes a program for identifying (estimating) the ink physical properties, ejection characteristics and / or print quality based on the electrical signals corresponding to the output operation of the drive signal.

[0018] The database 122 stores and retains information about the ink to be ejected. The ink referred to here may include various liquid materials, such as coloring materials for recording images, insulators for forming coatings and insulating layers, conductors for forming electrodes and electrical wiring, and semiconductors for forming electronic components. The information may include waveform information and environmental information of electrical signals detected in the signal path during and / or after the ink ejection operation, as well as the ink physical properties, ejection characteristics, and print quality of the ejected ink. These pieces of information are associated with each piece of waveform information. The database 122 may also include analysis results performed using this information and a machine learning model M trained on. The database 122 does not need to be physically located within the inkjet recording apparatus 1. For example, the database 122 may be located in an external auxiliary storage device, a network drive accessed via a network, a cloud server, or the like. Considering the possibility that the amount of data in the database 122 may increase significantly, the data may be distributed and stored across multiple storage devices.

[0019] The communication unit 13 has a network card and controls communication with the outside via a LAN (Local Area Network) or the Internet. The outside includes, for example, a computer (PC) that transmits print target data such as image data. The communication unit 13 may also be capable of communicating with an external server that stores a list of ink characteristic information and the like.

[0020] The display unit 14 has a digital display screen that displays settings and status of the recording operation under the control of the control unit 11. The digital display screen may be, for example, a liquid crystal display screen or an organic EL (Electro-Luminescent) screen.

[0021] The operation reception unit 15 receives an input operation from outside and outputs an operation signal corresponding to the received operation to the control unit 11. The operation reception unit 15 may have, for example, a push button switch, a slide switch, a rocker switch, a touch panel, etc. The touch panel may be positioned so as to overlap the digital display screen of the display unit 14.

[0022] The measurement unit 16 measures the environmental conditions of specific parts of the inkjet recording apparatus 1 and the surroundings. The measurement unit 16 may measure, for example, the ambient temperature and humidity, the temperature of the nozzle surface on which the nozzles 211 (see FIG. 2) that eject ink are arranged, and the temperature of the ink immediately before it is supplied to the nozzles 211. The measurement results are output to the control unit 11 at predetermined time intervals or at appropriate timing in response to a request.

[0023] The recording operation unit 20 performs a recording operation on a medium. In addition to a mechanism for ejecting ink, the recording operation unit 20 may also have mechanisms for supplying, transporting, fixing, and discharging the medium on which the ejected ink is applied and printed, and a mechanism for fixing the ink on the medium.

[0024] FIG. 2 is a diagram illustrating the recording operation unit 20. As shown in FIG. The recording operation unit 20 includes a recording element 21, a signal generating unit 22, a driving unit 23, a waveform information acquiring unit 24, an ink supply unit 25 (liquid material supply unit), and the like.

[0025] The recording element 21 includes a nozzle 211 that ejects ink, a pressure chamber 212 that communicates with the nozzle 211, and a piezoelectric element 213 that applies pressure fluctuations to the ink in the pressure chamber 212. The piezoelectric element 213 is an electromechanical transducer that is displaced (deformed) when a voltage is applied. The piezoelectric element 213 is in contact with the wall of the pressure chamber 212 directly or via a vibration plate or the like. The pressure of the ink in the pressure chamber 212 fluctuates as the wall of the pressure chamber 212 deforms in response to the deformation of the piezoelectric element 213. The deformation may be, for example, shear deformation or bending deformation, but is not limited to these. When the piezoelectric element 213 deforms in an appropriate pattern, ink with fluctuating pressure is ejected from the nozzle 211. Ink is supplied to the pressure chamber 212 from an ink supply unit 25 via an ink flow path. The inkjet recording device 1 may include multiple recording elements 21. The multiple recording elements 21 may be arranged in a predetermined pattern on one or more inkjet heads.

[0026] The ink supply unit 25 has an ink tank and an ink flow path. The ink tank stores the ink to be ejected. The ink flow path supplies ink from the ink tank to the pressure chambers 212. There may also be an ink flow path that collects ink from near the nozzles 211 and returns it to the ink tank. The ink supply unit 25 may be capable of supplying different types of ink with different characteristics to each of a predetermined number of recording elements 21 or inkjet heads.

[0027] The signal generating unit 22 outputs to the driving unit 23 a driving waveform that defines the amount and timing of change in the signal according to the driving signal to be output to the piezoelectric element 213. The signal generating unit 22 may have an oscillation circuit, or may be capable of converting an externally input clock signal into a set frequency and outputting it. Alternatively, the signal generating unit 22 may output a digital signal according to the driving waveform at a set cycle.

[0028] The driver 23 can adjust ink ejection parameters, such as whether or not to eject ink and / or the ejection amount, of each recording element 21 in accordance with drive data, such as image data, input from the storage unit 12. The driver 23 generates a drive signal that determines the ink ejection conditions in accordance with the adjustment result. The driver 23 outputs the generated drive signal to the piezoelectric element 213. When a digital signal of a drive waveform is input from the signal generating unit 22, the driver 23 may generate the drive signal by converting the digital signal to analog. The driver 23 can also appropriately amplify and output the generated drive signal. Note that the drive signals generated by the driver 23 are not limited to two types: an ejection signal that ejects ink and a signal that does not eject ink. The driver 23 may also be capable of generating multiple types of ejection drive signals according to the amount of ejected liquid. In addition, the signal that does not eject ink may be divided into a non-ejection signal that vibrates the ink in the pressure chamber 212 or the nozzle 211 and a no-operation signal that does not vibrate the ink. That is, in particular, when a signal for ejecting ink is not output during the image recording operation, a non-ejection signal or a no-operation signal is output during the same period.

[0029] The waveform information acquisition unit 24 detects an electrical signal in a signal path, which is a path electrically connecting the drive unit 23 and the piezoelectric element 213, and acquires the detected signal as waveform information. For example, the waveform information acquisition unit 24 includes a voltage sensor. The waveform information acquisition unit 24 may also include a control unit that determines a feature of a voltage change and includes the detected feature in the waveform information. The waveform information acquisition unit 24 outputs the obtained waveform information to the control unit 11. The waveform information acquisition unit 24 is included in the information acquisition unit of this embodiment. If the waveform information output by the waveform information acquisition unit 24 does not include a feature, the control unit 11 may determine a feature of a voltage change as part of the information acquisition unit of this embodiment and add the feature to the waveform information. While FIG. 2 shows a simple branching of a signal line from the signal path to the waveform information acquisition unit 24, the impedance and other parameters are determined to minimize current leakage to the waveform information acquisition unit 24. The circuit may also be configured so that the waveform of the electrical signal is not affected or changed by the presence of the waveform information acquisition unit 24 itself. Furthermore, if the waveform information acquisition unit 24 is not required during normal printing, a switching element or the like that blocks the transmission of signals to the waveform information acquisition unit 24 may be located between the branch and the waveform information acquisition unit 24. Alternatively, as long as waveform information can be acquired, the signal path does not need to physically have a branch path to the waveform information acquisition unit 24. Examples of such a configuration include a directional coupler and an isolator.

[0030] The ejection state observing unit 31 photographs or senses the state of ink droplets ejected from the nozzles 211, and outputs the observed results as information on ejection characteristics to the control unit 11. The inkjet recording device 1 may also be equipped with an imaging unit (not shown) or the like, which can capture an image of the printing surface of the printed medium and output the image as information on print quality to the control unit 11. The ejection state observing unit 31 may be a separate device from the inkjet recording device 1, which is attached as needed.

[0031] The control unit 11 can separately acquire environmental information from the measurement unit 16. The control unit 11 can also acquire information about ink physical properties, ejection characteristics, and / or the quality of printed matter (print quality) via the communication unit 13 or the operation reception unit 15. The ink physical property information may be collected and stored in an external server device or the like as part of information about multiple inks. The ejection characteristic information may be obtained by the control unit 11 analyzing and evaluating the state of ink droplets acquired by the ejection state observation unit 31. The print quality information may be obtained by the control unit 11 analyzing and evaluating the image of the printed matter captured by an imaging unit (not shown).

[0032] When generating the database 122, the control unit 11 acquires the above information, associates the information with each other, and writes and stores the information in the database 122. When using the database 122, the control unit 11 accesses the database 122 based on the acquired waveform information, environmental information, and the like.

[0033] Next, the waveform information will be described. As described above, the waveform of the drive signal output by the drive unit 23 defines the deformation pattern of the piezoelectric element 213. The pressure chamber 212 is compressed by the deformation of the piezoelectric element 213, thereby increasing the ink pressure in the pressure chamber 212. The pressure chamber 212 is expanded by the deformation of the piezoelectric element 213, thereby decreasing the ink pressure in the pressure chamber 212.

[0034] The drive waveform may be generated by combining multiple frequency components. The multiple frequency components may be multiplexed on the frequency axis of the drive waveform. The drive waveform may have, for example, a rectangular or trapezoidal pulse shape, or may have a binary signal shape or a multi-value signal shape. The drive waveform for one ink ejection may include multiple rectangular waves, trapezoidal waves, or a combination thereof. At least one of the multiple rectangular waves or trapezoidal waves may have a different shape from the others. Some of the multiple rectangular waves or trapezoidal waves may change in a different direction (positive or negative) from the reference voltage than the remaining part.

[0035] The frequency components that can be included in such a drive waveform are set to suitably match the resonance frequency associated with the pressure fluctuation of the ink inside the pressure chamber 212 and the resonance frequency associated with the movement of the ink liquid surface in the nozzle 211. This allows the ink to efficiently protrude from the nozzle 211 and the protruding ink to be separated from the ink inside the nozzle 211.

[0036] On the other hand, the piezoelectric element 213, which is an electromechanical conversion element, receives pressure from the ink. That is, the piezoelectric element 213 generates an electric signal as the piezoelectric element 213 deforms in response to fluctuations in ink pressure. That is, the electric signal actually detected in the signal path between the piezoelectric element 213 and the driver 23 includes a component of the drive signal generated by the driver 23 and to be output to the piezoelectric element 213, and a component of the electric signal corresponding to deformation due to ink pressure acting on the piezoelectric element 213. For example, the signal waveform is a composite waveform of both signals. Therefore, the electric signal detected in the signal path connected to the piezoelectric element 213 is accompanied by a waveform dullness or distortion compared to the original drive signal in response to the pressure from the ink.

[0037] The dynamic pressure exerted by the ink, which moves or compresses in response to pressure fluctuations and ejection, can be affected by various characteristics (physical properties) of the ink. At the same time, the ink ejection state in response to the deformation of the piezoelectric element 213 also depends on various physical properties of the ink. In other words, the ink ejection operation is a complex physics involving ink, ink flow paths, and piezoelectric elements.

[0038] In this embodiment, as described above, waveform information of an electrical signal actually detected by the piezoelectric element 213, ink physical property information, ejection characteristic information, print quality information, and the like are associated and stored in advance as the database 122. As described above, the association may also include environmental information during ink ejection operations. The relationships between these are learned based on this database 122. For any of the ink physical property information, ejection characteristic information, and print quality information, the waveform information and / or various feature quantities of the electrical signal detected by the piezoelectric element 213 are treated as multiple variables. Then, based on the results of this learning, the ink physical properties, ejection characteristics, and print quality are estimated for the waveform information of the actual electrical signal. The analysis results may be stored outside the database 122.

[0039] As described above, the database contains as many parameter variables as possible related to the final print quality and the ink ejection operation that can affect the print quality. The parameters include the ink physical properties, ejection characteristics, environmental information that can affect the ink physical properties, ejection characteristics, and print quality, and information on the actual electrical signal detected by the piezoelectric element 213.

[0040] On the other hand, the characteristics of the nozzles 211, pressure chambers 212, and ink flow paths, such as their shapes and sizes, as well as the characteristics of the piezoelectric elements 213, are determined for each inkjet recording apparatus 1. While these are constants, the information on the electrical signals that depend on these constants and the final print quality are meaningless unless they are results obtained under these constants. Therefore, the database is generated for each inkjet recording apparatus 1 within that inkjet recording apparatus 1.

[0041] The print quality may include some or all of image bleeding, resolution, glossiness, density gradation value, and color tone (brightness, saturation, hue, etc.), for example.

[0042] The physical properties of the ink may include some or all of the following: viscosity, modulus of elasticity, density, specific gravity, concentration of each component (solid component), dispersed particle size of the pigment, particle size distribution, surface tension, thixotropy, gas content, dielectric constant, charge rate, refractive index, zeta potential, etc.

[0043] The ejection characteristics may include some or all of the ejection speed, the amount, size and shape of the ejected ink droplets, the accuracy of the droplet landing position (straightness), the number of satellites, and the variation and stability of these parameters.

[0044] The environmental information may include, for example, at least one of temperature and humidity.

[0045] Waveform information of the electrical signal includes, for example, voltage amplitude, rise and fall times of the voltage change, overshoot and undershoot amounts, distortion rate, and harmonic frequency and amount. The rise and fall times include not only the time corresponding to the signal waveform output by the signal generating unit 22 but also delay times resulting from the characteristics of the electrical circuitry of the inkjet recording apparatus 1. The overshoot and undershoot amounts represent the magnitude of the voltage change that exceeds the original voltage level (set voltage) set in the drive signal during the sudden voltage change, such as the rectangular or trapezoidal shape. In addition, ink may be thixotropic. That is, the physical properties of the ink may depend on factors such as the duration of operation of the piezoelectric element 213. In the case of such ink or ink that may be thixotropic, the waveform information of the electrical signal may also include values ​​corresponding to the duration of operation of the piezoelectric element 213 or the interval since the previous operation.

[0046] The waveform information of the electrical signal may include the distribution shape of the frequency spectrum and the waveform shape itself. These may be stored as image data as well as numerical data. Furthermore, the waveform information of the electrical signal may include waveform information of the original drive signal and a difference from the feature amount of the original drive signal.

[0047] FIG. 3 is a flowchart showing the control procedure of the database generation process executed by the inkjet recording apparatus 1.

[0048] The control unit 11 sets a known ink as a target for ejection. The control unit 11 acquires physical property information of the set known ink (S1). The ink physical property information may be acquired, for example, from an external server based on the ink's identification information. In addition to or instead of this, the ink physical property information may be acquired by an input operation to the operation reception unit 15. The known ink is also set in the ink supply unit 25 by a separate manual operation by a person in charge. Note that, in the process S2 below and subsequent processes, a certain known ink may be ejected multiple times depending on the same or different environments. Therefore, when a known ink is set multiple times in succession, the ink physical property information does not need to be acquired again from an external device, the operation reception unit 15, etc.

[0049] The control unit 11 acquires environmental information from the measurement unit 16 (S2). As described above, the environmental information may include temperature and humidity.

[0050] The control unit 11 controls the ink supply unit 25 to supply the selected ink to the recording element 21 via the ink supply path (S3). The control unit 11 controls the signal generating unit 22 to generate a predetermined drive waveform and the driving unit 23 to output a drive signal (first drive signal) related to the ink ejection operation toward the piezoelectric element 213 in accordance with the drive waveform (S4). The control unit 11 controls the waveform information acquiring unit 24 to acquire waveform information (first waveform information) of an electric signal (first electric signal) actually detected in the signal path between the driving unit 23 and the piezoelectric element 213 (S5). The waveform information of the electric signal is acquired, for example, during the period when the ink ejection operation is being performed in response to the drive signal. Additionally or alternatively, the waveform information of the electric signal may also be acquired for a predetermined period after the ink ejection operation in response to the drive signal has ended. For example, depending on the physical properties of the ink and the structure of the recording element 21, a voltage fluctuation component may remain and be detected between the driving unit 23 and the piezoelectric element 213 even after the ink has been ejected. This may be acquired as waveform information of the electric signal. As described above, the control unit 11 may calculate the feature amount of the electrical signal acquired by the waveform information acquisition unit 24.

[0051] The control unit 11 acquires ink ejection characteristic information (S6). The ejection characteristic information may be acquired by the control unit 11 analyzing the ejection characteristics acquired from the ejection state observing unit 31. Alternatively, the control unit 11 may acquire the ejection characteristic information from an external source via the communication unit 13 or the operation accepting unit 15.

[0052] Control unit 11 acquires print quality information (S7). The print quality information may be acquired by control unit 11 analyzing the captured image data of the print image, as described above. Alternatively, the captured image data of the print image may be analyzed by an external device, and the analysis results may be acquired from the external device via communication unit 13. Additionally or alternatively, the print quality information may be acquired by operation acceptance unit 15 receiving the results of a human evaluation of the print image and quantification thereof. As will be described later, when the object to be estimated based on the database 122 is ink physical properties, ejection characteristic information and print quality information are not necessarily required. In this case, steps S6 and S7 may be omitted. Also, when the object to be acquired based on the database 122 is ejection characteristics, ink physical property information and print quality information are not necessarily required. In this case, in step S1, the control unit 11 does not need to acquire the set known ink physical property information, and step S7 may be omitted. Also, when the object to be acquired based on the database 122 is abnormal print quality, ink physical property information and ejection characteristic information are not necessarily required. In this case, in step S1, the control unit 11 does not need to acquire the set known ink physical property information, and step S6 may be omitted.

[0053] The control unit 11, as a database generating unit, associates the obtained waveform information, environmental information, ink physical property information, ejection characteristic information, and print quality information, and stores them in the database 122 (S8).

[0054] The control unit 11 determines whether all data to be acquired has been acquired and the generation of the database is to be terminated (S9). The data to be acquired may be simply determined by the number of data to be acquired, or the number of times to acquire each of the ink types prepared in advance may be determined. If it is determined not to terminate the generation of the database (S9; N), the processing of the control unit 11 returns to step S1. If it is determined to terminate the generation of the database (S9; Y), the control unit 11 performs an analysis process of the data stored in the database 122 (S10). The analysis process may include learning of the machine learning model M. The input of this machine learning model M includes at least waveform information. The output of the machine learning model M is ink physical properties, ejection characteristics, or print quality. The control unit 11 terminates the database generation control process.

[0055] The analytical process executed in step S9 using the generated database 122 will now be described. In a database with a large number of parameters, it is difficult to theoretically or deductively determine the relationship between a certain input parameter and another certain output parameter, as described above. Therefore, the relationship is analyzed and derived by numerical processing. The method used for the analysis is not particularly limited. Commonly used analytical methods, such as cluster diffraction and association diffraction, may be used. Alternatively, other appropriate analytical methods may be employed, or multiple analytical methods may be combined.

[0056] The machine learning model M used in such analysis processing is assumed to be configured to output an estimate of a certain output parameter for a large number of input parameters (multidimensional data), as described above. For example, a support vector machine, a decision tree, or ensemble learning related thereto may be used for such a machine learning model M. Furthermore, a configuration using deep learning, such as a convolutional neural network, may be used for image recognition, feature extraction, and the like. The error is fed back to the parameters of the machine learning model M so that the estimate output for the input of the multidimensional data approaches the estimate stored in the database 122. For example, backpropagation may be used for the feedback.

[0057] As described above, the parameters of the data in database 122 that are to be output by the relational expressions and machine learning model M during the estimation process may be the final goal of print quality, ink physical properties, or ejection characteristics. The input parameters may be all parameters other than the output parameters registered in the database. In particular, the input parameters include waveform information related to electrical signals. However, some of the parameters other than the output parameters may be excluded from the input. For example, parameters that are not sufficiently numerous for training the machine learning model M, or parameters that are determined to be irrelevant through multiple analyses, may be excluded.

[0058] In this way, the relationships between many parameter variables and the output target parameters are determined. Such database 122 and the analysis results may be updated as frequently as appropriate. That is, the database generation control process may be executed multiple times in response to an increase or update of data used in the analysis.

[0059] Conversely, the analysis process may involve determining correlations between ink physical property information, ejection characteristic information, and / or print quality information as input parameters, with the goal of outputting waveform information. Furthermore, analysis may be performed so that the input also includes target values ​​for improvement, such as the degree to which each parameter of the ink physical property information, ejection characteristic information, and print quality information should be improved. As described below, this analysis can provide waveform information corresponding to the desired ink physical properties, ejection characteristics, and print quality. Furthermore, as described above, there is a difference between the drive signal for ink ejection and the electrical signal actually output from the drive unit 23 and applied to the piezoelectric element 213. Therefore, analysis may be performed using the output parameters as the drive signal.

[0060] Based on the database 122 obtained in the above manner, estimation processing is performed for the ink to be evaluated.

[0061] FIG. 4 is a flowchart showing the control procedure for ink property estimation processing using the above analysis results. The control unit 11 acquires environmental information from the measurement unit 16 (S21). With the ink to be evaluated set in the ink supply unit 25, the control unit 11 causes the ink supply unit 25 to supply ink to the inkjet head (S22).

[0062] The control unit 11 causes the signal generating unit 22 to generate a predetermined drive waveform and causes the drive unit 23 to output a drive signal (second drive signal) corresponding to the drive waveform (S23). This drive signal has waveform parameters such as standard amplitude and period for ejecting ink. These waveform parameters may be the same as some or all of the waveform parameters of the drive signal output in step S4 of the database generation process. This ejects the ink to be evaluated. Steps S22 and S23 correspond to the processing of the control unit 11 as an operation control unit in this embodiment. The control unit 11, as an information acquisition unit, causes the waveform information acquisition unit 24 to acquire waveform information (second waveform information) of the electrical signal (second electrical signal) in the signal path of the piezoelectric element 213 (S24; acquisition means).

[0063] The control unit 11, functioning as an estimation unit, extracts ink physical properties that have a plausible correspondence with the environmental information and waveform information based on the database 122. The control unit 11 estimates the ink physical properties of the ink to be evaluated based on the extraction results (S25; estimation means). "Based on the database 122" includes cases where the ink physical properties are obtained using a relational expression or machine learning model M obtained as a result of analyzing the database 122, as described above. Therefore, if the relational expression or machine learning model M is located outside the database 122, estimation based on the database 122 does not directly access the data stored in the database 122. The extracted information on the ink physical properties is output as an estimation result for the ink to be evaluated via the display unit 14, the communication unit 13, or the like. The control unit 11 then terminates the ink physical property estimation process.

[0064] The obtained estimation results of the ink physical properties may be used, for example, to develop improved ink.

[0065] FIG. 5 is a flowchart showing a control procedure for the ejection characteristics estimation process using the analysis results. The ejection characteristics estimation process has the same contents as the ink physical properties estimation process described above in steps S21 to S24. Therefore, detailed descriptions of these processes will be omitted. However, if some or all of the ink physical properties to be evaluated are known, the control unit 11 may acquire these in step S21. After step S24, the control unit 11, functioning as an estimation unit, estimates the ejection characteristics based on the database 122 (S25a; estimation means). If information on the known ink physical properties has been acquired as described above, this information can also be added to the input parameters for estimation based on the database 122.

[0066] As with the ink physical property estimation process described above, "based on the database 122" does not necessarily mean direct access to the database 122. When a relational expression based on the results of the analysis of the database 122 or a machine learning model M is located outside the database 122, the results obtained from these may be acquired. The obtained ejection characteristic information is output as an estimation result of the ejection characteristics via the display unit 14, the communication unit 13, or the like. Then, the control unit 11 ends the ejection characteristic estimation process.

[0067] The obtained ejection characteristics may be used to improve the ink or the piezoelectric element.

[0068] FIG. 6 is a flowchart showing the control procedure for the print quality estimation process using the analysis results. The print quality estimation process has the same contents as the ink physical properties estimation process described above in steps S21 to S24. Therefore, detailed descriptions of these processes will be omitted. However, if some or all of the ink physical properties and / or ejection characteristics to be evaluated are known, the control unit 11 may acquire these in step S21. After step S24, the control unit 11, as an estimation unit, estimates the print quality based on the database 122 (S25b; estimation means). If information on the known ink physical properties and / or ejection characteristics has been acquired as described above, this information can also be added to the input parameters of the estimation based on the database 122.

[0069] As with the ink physical property estimation process described above, "based on database 122" does not necessarily mean direct access to database 122. When a relational expression based on the results of an analysis of database 122 or machine learning model M is located outside database 122, the results obtained from these may be acquired. The obtained print quality information is output as the print quality estimation result via display unit 14, communication unit 13, or the like. Then, control unit 11 ends the print quality estimation process.

[0070] The resulting print quality may be utilized to improve inks and inkjet heads.

[0071] [Second embodiment] Next, an inkjet recording apparatus 1a according to a second embodiment will be described. FIG. 7 is a diagram illustrating the recording operation unit 20a of the inkjet recording apparatus 1a. The hardware configuration of the inkjet recording apparatus 1a other than the recording operation unit 20a is the same as that of the inkjet recording apparatus 1, and therefore will not be illustrated or described.

[0072] The recording operation unit 20a is capable of supplying inks having different characteristics to the plurality of recording elements 21a to 21c of the signal generating unit. The recording operation unit 20a is also capable of generating a plurality of drive waveforms, namely, a first waveform 221 to a third waveform 223, using the signal generating unit 22a. The recording operation unit 20a is capable of outputting a first signal 231 to a third signal 233, which are drive signals corresponding to the first waveform 221 to the third waveform 223, using the drive unit 23a. The recording operation unit 20a is also capable of acquiring waveform information of the electrical signals corresponding to the first signal 231 to the third signal 233 from the electrical connection paths between the drive unit 23a and the piezoelectric elements 213a to 213c using the waveform information acquiring unit 24. The drive signals that can be set as the first signal 231 to the third signal 233 may include, as described above, an ejection signal, a non-ejection signal, a no-operation signal, and the like.

[0073] There may be a plurality of recording elements 21a to which the first signal 231 is input. There may be a plurality of recording elements 21b to which the second signal 232 is input. There may also be a plurality of recording elements 21c to which the third signal 233 is input. The plurality of recording elements 21a may belong to one or more of the plurality of inkjet heads. The plurality of recording elements 21b may belong to one or more inkjet heads different from the plurality of recording elements 21a. The plurality of recording elements 21c may belong to one or more inkjet heads different from the plurality of recording elements 21a, 21b.

[0074] Depending on the situation, the first waveform 221 to the third waveform 223 may be partly or entirely the same, or the first waveform 221 to the third waveform 223 may all be different from one another. Accordingly, the first signal 231 to the third signal 233 may be partly or entirely the same.

[0075] As described above, the recording elements 21a to 21c may be divided to eject different types of ink, or the recording elements 21a to 21c that normally eject a common ink may be switched to eject different types of ink when generating the database 122.

[0076] 7 shows a configuration in which there are three sets of drive waveforms, drive signals, and electrical signals corresponding to the recording elements, but the number may be two or more. That is, the number of recording elements, the number of drive waveforms that the signal generating unit 22a can output in parallel, and the number of electrical signals that the driving unit 23a can output in parallel and detect in the signal path to the piezoelectric elements may each be determined as appropriate. This configuration allows data to be stored in database 122 in parallel for known inks having at least two different characteristics and / or waveform information for electrical signals associated with at least two types of drive signals. This improves acquisition efficiency, particularly when acquiring data combining numerous ink properties and electrical signals.

[0077] [Third embodiment] Next, an inkjet recording apparatus 1b according to a third embodiment will be described. FIG. 8 is a diagram illustrating a recording operation unit 20b of the inkjet recording apparatus 1b.

[0078] In the inkjet recording device 1b, the recording operation unit 20a of the inkjet recording device 1a is replaced with a recording operation unit 20b. The recording operation unit 20b has a transfer flow path 26. The transfer flow path 26 sends ink ejected from a certain nozzle 211 to the ink flow path of another nozzle 211. Here, pressure fluctuations are applied to the ink in the pressure chamber 212a in response to the displacement of the piezoelectric element 213a (first electromechanical conversion element) in the recording element 21a (first recording element). In response to this, the ink ejected from the nozzle 211a flows into the transfer flow path 26. The ink that has flowed into the transfer flow path 26 is guided to the ink flow path of the recording element 21b (second recording element) having another piezoelectric element 213b (second electromechanical conversion element), and then flows into the pressure chamber 212b. The length and shape of the transfer flow path 26, particularly the bends, may be determined so as to minimize the loss of kinetic energy of the ink ejected from the nozzle 211a. Furthermore, the positional relationship between the inkjet head including the recording element 21a and the inkjet head including the recording element 21b may be adjustable so as to reduce the loss of kinetic energy.

[0079] The signal generating unit 22a and the driving unit 23a can output different driving waveforms and driving signals to the piezoelectric elements 213a and 213b, respectively. The driving unit 23a outputs a driving signal (ejection signal) to the recording element 21a to eject ink, and does not output a driving signal to the recording element 21b while the transfer channel 26 is connected. Alternatively, the driving unit 23a may output a non-operation signal that does not impart pressure fluctuations to the ink as the second signal 232 to the recording element 21b. Ink ejected from the nozzle 211a of the recording element 21a applies pressure to the piezoelectric element 213b in the pressure chamber 212b of the recording element 21b. The waveform information acquiring unit 24 detects an electrical signal that is not directly affected by the output operation of the driving signal by the driving unit 23a in the signal path leading to the piezoelectric element 213b, and can acquire high-quality waveform information. In this case, it is not necessary for ink to be ejected from nozzle 211b, and print quality information corresponding to ink that has actually been ejected does not necessarily have to be acquired.

[0080] [Fourth embodiment] Next, an inkjet recording apparatus 1c according to a fourth embodiment will be described. FIG. 9 is a diagram illustrating a recording operation unit 20c of the inkjet recording apparatus 1c.

[0081] In the inkjet recording apparatus 1c, the recording operation unit 20b of the inkjet recording apparatus 1b is replaced with a recording operation unit 20c. The recording operation unit 20c generates a differential signal between two electric signals. One of the two electric signals is an electric signal detected between the drive unit 23a and the piezoelectric element 213a (third electromechanical conversion element) of the recording element 21a (third recording element). The other of the two electric signals is an electric signal detected between the drive unit 23a and the piezoelectric element 213b (fourth electromechanical conversion element) of the recording element 21b (fourth recording element). As the differential processing for generating this differential signal, in FIG. 9, the electric signal detected from the signal path connected to the piezoelectric element 213a is inverted in sign and added to the electric signal detected from the signal path connected to the piezoelectric element 213b, but the present invention is not limited to this.

[0082] As in FIG. 2, the ink supply unit 25 supplies ink from an ink tank to at least the pressure chamber 212a via an ink flow.

[0083] The signal generating unit 22a and the driving unit 23a can output different driving waveforms and driving signals to the piezoelectric elements 213a and 213b, respectively. The driving unit 23a outputs an ejection signal to the recording element 21a as the first signal 231, which causes ink to be ejected, and does not output a driving signal to the recording element 21b. Alternatively, the driving unit 23a may output a non-operation signal as the second signal 232, which does not impart pressure fluctuations to the ink. In this state, the waveform information acquiring unit 24 acquires waveform information of the differential signal. The piezoelectric elements 213a and 213b may be spatially close to each other. If noise associated with the output of the first signal 231 and the operation of other components is detected as an electrical signal from the signal path connected to the piezoelectric element 213b, the differential signal can more accurately include the influence of deformation of the piezoelectric element 213a, thereby acquiring high-quality waveform information that reflects the physical properties of the ink.

[0084] Although the recording operation unit 20c has been described as including a circuit for calculating the difference in hardware, this is not limiting. If the waveform information acquisition unit 24 can detect electrical signals with sufficient time resolution, the waveform information acquisition unit 24 may calculate the difference between two digital signals obtained in software. In this case, the piezoelectric element 213 that detects the two electrical signals for which the difference is calculated can be easily changed.

[0085] [Fifth embodiment] Next, an inkjet recording apparatus 1d according to a fifth embodiment will be described. FIG. 10 is a diagram illustrating a recording operation unit 20d of the inkjet recording apparatus 1d.

[0086] The recording operation unit 20d can change the frequency of the clock signal generated by the signal generating unit 22d and the frequency component of the drive waveform in response to a signal from the control unit 11. As described above, the resonant frequency of the ink pressure vibration and the resonant frequency of the liquid surface within the nozzle 211 are generally determined depending on the structure of the inkjet head. Therefore, the frequency of the drive waveform generated by the signal generating unit 22d does not need to be changed during normal ink ejection operations.

[0087] In contrast, in this embodiment, the frequency of the drive waveform generated by the signal generating unit 22d is variable. This allows the drive unit 23 to output drive signals with different frequency components. Accordingly, the database 122 can store waveform information for electrical signals obtained when ink ejection operations are performed using drive signals with the above-mentioned different frequency components. Such drive signals may intentionally include signals with frequency components that are not suitable for ink ejection. The frequency components that are not suitable for ink ejection may include frequency components lower than the frequency components of drive signals used for normal ink ejection. The drive signal may have one or more frequency components. Multiple patterns of settable drive waveforms may be determined in advance.

[0088] The drive signal may have multiple frequency components, and the multiple frequency components may be multiplexed on the frequency axis of the drive signal or on the time axis, i.e., different frequency components may be output sequentially in time series.

[0089] In one embodiment, the drive signal with a frequency component that is not suitable for ink ejection may include, for example, a frequency (of the order of several MHz) that corresponds to the shear speed at which the ink is fixed on the medium onto which it has been ejected. Ink ejection operations based on such drive signals do not necessarily involve actual ejection of ink from the nozzles 211. Accordingly, ejection characteristic information and print quality information do not necessarily need to be acquired.

[0090] In this way, by using a drive signal with a different frequency component from the drive signal for actual ink ejection, waveform information of the different frequency and the corresponding ejection characteristic information and print quality information can be obtained. By storing ink physical property information corresponding to the different frequency components and this information in database 122, it is possible to estimate ink physical properties, ejection characteristics, or print quality from a wider range of evaluation items, such as ink stability on a medium.

[0091] FIG. 11 is a flowchart showing the control procedure for database generation processing in the inkjet recording apparatus 1d of this embodiment. In this database generation control process, process S9 is deleted and processes S11 to S13 are added instead to the database generation control process shown in Fig. 3. The other processes are the same, and the same process contents are assigned the same reference numerals and detailed explanations are omitted.

[0092] After step S1, the control unit 11 sets the frequency of the drive waveform generated by the signal generating unit 22d (S11). The process of the control unit 11 proceeds to step S2.

[0093] After step S8, the control unit 11 determines whether data has been acquired at all frequencies for the set ink (S12). If it is determined that data has not been acquired at all frequencies (S12; N), the control unit 11 returns to step S11.

[0094] If it is determined that data has been acquired for all frequencies (S12; Y), the control unit 11 determines whether data has been acquired for all target ink types (S13).

[0095] If it is determined that data has not been acquired for all ink types (S13; N), the control unit 11 returns to step S1. If it is determined that data has been acquired for all ink types (S13; Y), the control unit 11 proceeds to step S9.

[0096] [Sixth embodiment] Next, an inkjet recording apparatus 1e according to a sixth embodiment will be described. FIG. 12 is a diagram illustrating a recording operation unit 20e of the inkjet recording apparatus 1e. In this embodiment, the control unit 11 can change and set the drive waveform generated by the signal generating unit 22e or the drive signal output by the drive unit 23e based on the database 122 so as to obtain the desired ejection characteristics or print quality.

[0097] The control unit 11 extracts waveform information corresponding to desired ejection characteristics or print quality information based on the database 122. The control unit 11 controls the driver 23e to adjust the set drive signal based on the difference between the current waveform information and the extracted waveform information. Alternatively, the control unit 11 may control the signal generating unit 22e to adjust the set drive waveform, thereby controlling the driver 23e to adjust the set drive signal. By outputting a drive signal that has been suitably adjusted in this way from the driver 23e, the ejection characteristics or print quality can be brought closer to the desired level.

[0098] In the above description, the control unit 11 performs control operations based on the relationship between waveform information and ejection characteristics or print quality. However, the control operations may also take into account environmental information and the ink properties when known inks are used. In other words, it is sufficient to adjust the drive signal appropriately according to the ink properties of the ink used and environmental information so that waveform information corresponding to the desired waveform information can be obtained from the electrical signal.

[0099] FIG. 13 is a flowchart showing the control procedure of the print quality adjustment process in one embodiment. Before this process is executed, the process of generating the database 122 is completed based on the flowchart shown in FIG.

[0100] In this process, steps S21 and S23 are the same as the processes described in the ink physical property estimation process shown in Fig. 4. Therefore, detailed description of these processes will be omitted.

[0101] After step S21, control unit 11 acquires desired print quality information (target) (S41). The acquired print quality information may have been selected and set in advance based on selection information displayed on display unit 14. Alternatively, data on print quality selection information set externally or the like may be stored and held in memory unit 12 via communication unit 13, and this information may be acquired.

[0102] The control unit 11 executes the print quality estimation process shown in FIG. 6 (S42). As described above, the phrase "based on the database 122" does not necessarily mean direct access to the database 122. It also includes the use of a multivariate function (relational expression) or a machine learning model M obtained by analyzing the database 122. Furthermore, the input parameters may include known ink physical property information, ejection characteristic information, and environmental information, as described above. The control unit 11, as an adjustment unit, adjusts the drive signal by setting a drive waveform based on a comparison between the estimated print quality and the desired print quality (S43). This allows the electrical signal actually detected by the piezoelectric element 213 to be adjusted appropriately for the desired print quality. Instead of the estimated print quality, the print quality may be obtained by actually ejecting ink and printing, and then compared with the desired print quality. In this case, the print quality estimation process (S42) may be omitted. Furthermore, if information about the drive waveform to be set can be obtained solely from the desired print quality, process S42 may also be omitted.

[0103] The control unit 11 causes the ink supply unit 25 to supply ink to the inkjet head (S22a). The control unit 11 causes the drive unit 23 to output an electrical signal related to ink ejection based on the data to be printed (S23). As a result, the control unit 11 performs printing on the medium and forms a printed matter. The control unit 11 then ends the print quality adjustment process.

[0104] FIG. 14 is a flowchart showing a control procedure for print quality adjustment processing in another embodiment. This print quality adjustment process is the same as the print quality adjustment process of Fig. 13, except that process S44 has been added. The same process contents are given the same reference numerals and detailed explanations will be omitted.

[0105] After step S43, the control unit 11 further sets a frequency suitable for ink ejection (S44), and then the process of the control unit 11 proceeds to step S22a.

[0106] As described above, the inkjet recording apparatus 1 according to the evaluation method of this embodiment includes a drive unit 23 and a recording element 21. The recording element 21 includes a piezoelectric element 213. A drive signal is output from the drive unit 23 to displace the piezoelectric element 213, thereby ejecting ink and applying the ink to a medium to perform printing. This evaluation method involves acquiring information about the ink ejected by the inkjet recording apparatus 1 and the printed image formed by the ink. This evaluation method estimates at least one of the ink's physical properties, ejection characteristics, and print quality based on waveform information of an electrical signal detected at the input side of the drive signal to the piezoelectric element 213. The waveform information of the electrical signal actually applied to the piezoelectric element 213 includes the influence of the force applied to the piezoelectric element 213 by a liquid material, such as ink. In particular, in situations where the liquid material dynamically moves or compresses in association with the ink ejection operation, this waveform information contains a large amount of information about the liquid material. Therefore, this evaluation method estimates at least one of the ink's physical properties, ejection characteristics, and print quality based on the waveform information. As a result, according to this evaluation method, more appropriate information about the ink ejected from the inkjet recording apparatus 1 can be easily obtained.

[0107] Furthermore, in this evaluation method, a database 122 may be generated that relates waveform information obtained when printing is performed by ejecting a known ink to at least one of the ink physical properties, ejection characteristics, and print quality of the known ink. The above estimation may also be performed using this database 122. By performing such estimation, it becomes possible to easily obtain more appropriate information about the ink more easily and accurately according to the correspondence determined from a large amount of data.

[0108] The inkjet recording apparatus 1 may also have a plurality of recording elements 21. In this evaluation method, at least two types of known inks having different characteristics may be ejected from the plurality of recording elements 21, respectively, to perform printing using the at least two types of ink. A database 122 may be generated that relates waveform information to at least one of the ink physical properties, ejection characteristics, and print quality of the known inks corresponding to the plurality of recording elements 21 when performing this printing. By performing the above estimation based on such a database 122, the correspondence can be obtained more easily, making it easier to quickly obtain more appropriate information about the inks.

[0109] Furthermore, in this evaluation method, at least one of the ink properties, ejection characteristics, and print quality of the ink to be evaluated may be estimated based on waveform information obtained when the ink to be evaluated is ejected and printed, using database 122. Such estimation makes it easier to improve the ink used in inkjet recording apparatus 1 and the quality of printed matter produced by inkjet recording apparatus 1 with greater accuracy.

[0110] Alternatively, the waveform information may be obtained from the electrical signal actually output when the driver 23 is outputting a drive signal that causes the recording elements 21 to eject ink. The electrical signal obtained during the period in which the ink state dynamically changes in response to the drive waveform provides a wide variety of multivariate information corresponding to the ink's characteristics (physical properties and ejection characteristics). Furthermore, this electrical signal is more closely linked to print quality. Therefore, in this evaluation method, by using the electrical signal obtained by the inkjet recording device 1 during this period, information that more accurately reflects the ink's characteristics can be obtained.

[0111] Alternatively, the waveform information may be obtained from an electrical signal after the driver 23 outputs a drive signal that causes the recording element 21 to eject ink. In other words, estimated results of ink physical properties, ejection characteristics, and / or print quality may be obtained based on reverberation vibrations that occur in the piezoelectric element 213 in response to ink movement and pressure changes after the drive signal is output. In this way, in this evaluation method, by obtaining waveform information of an electrical signal that does not include components of the drive waveform, it is possible to more easily obtain information that reflects the ink characteristics.

[0112] Furthermore, the estimation may be performed using a machine learning model that outputs at least one of ink physical properties, ejection characteristics, and print quality in response to input waveform information. The machine learning model makes it possible to efficiently obtain estimated results for desired parameters based on multidimensional parameters, particularly in a range that cannot be achieved by manual work or human understanding.

[0113] Alternatively, the electrical signal may be detected from a path electrically connecting the driving unit 23 and the piezoelectric element 213. This makes it possible to easily obtain waveform information that is stored in a database and used for estimation.

[0114] The inkjet recording apparatus 1 may also have a plurality of recording elements 21. The recording operation unit 20b may include a transfer flow path 26 that guides ink ejected in response to the displacement of the piezoelectric element 213a of the recording element 21a to a recording element 21b different from the recording element 21a. The electrical signal may be detected from a path electrically connecting the driving unit 23a and the piezoelectric element 213b of the recording element 21b. In this case, information corresponding to the pressure change and movement of the ink accompanying the ink ejection operation of the piezoelectric element 213a due to the first signal 231 can be obtained from the electrical signal detected in the path electrically connecting the driving unit 23a and the piezoelectric element 213b. This makes it possible to obtain waveform information of higher quality that is stored in a database and used for estimation.

[0115] The driver 23a may be able to select whether or not to output a drive signal related to ink ejection to each of the piezoelectric elements 213a and 213b of the recording elements 21a and 21b. That is, the electrical signal may be detected when the driver 23a outputs a drive signal to the piezoelectric element 213a but does not output a drive signal to the piezoelectric element 213b. In this case, pressure fluctuations and movement occur in the ink in the pressure chamber 212b guided by the transfer channel 26, causing the piezoelectric element 213b to generate an electrical signal. That is, the electrical signal detected from the path electrically connecting the driver 23a and the piezoelectric element 213b does not include the drive signal output from the driver 23a. This allows for higher quality waveform information to be stored in the database and used for estimation.

[0116] The inkjet recording apparatus 1 may also have a plurality of recording elements 21. The waveform information may include the difference between the following two electrical signals: (1) an electrical signal detected from a path electrically connecting the driving unit 23a and the piezoelectric element 213a of the recording element 21a; and (2) an electrical signal detected from a path electrically connecting the driving unit 23a and the piezoelectric element 213b of the recording element 21b. From such differential signals, the inkjet recording apparatus 1 can obtain information such as the difference in the electrical signals between the recording elements 21a and 21b.

[0117] The driver 23a may be capable of selecting whether to output a drive signal related to ink ejection to each of the piezoelectric elements 213a and 213b of the recording elements 21a and 21b. That is, the electrical signal may be detected when the driver 23a outputs a drive signal to the piezoelectric element 213a but not to the piezoelectric element 213b. In this case, the ink in the pressure chamber 212b does not experience pressure fluctuations or movement, and the piezoelectric element 213b does not generate an electrical signal. Therefore, the electrical signal detected from the electrical path connecting the driver 23a and the piezoelectric element 213b is noise related to the operation of the driver 23a, the recording element 21a, the recording element 21b, or other components. In the waveform information, such noise may be subtracted from the electrical signal detected from the electrical path connecting the driver 23a and the piezoelectric element 213a. This allows for higher-quality waveform information to be stored in a database and used for estimation.

[0118] The ink physical properties may also include at least one of viscosity, elasticity, density, specific gravity, solids concentration, surface tension, thixotropy, gas content, dielectric constant, charge rate, refractive index, zeta potential, dispersed particle size, and particle size distribution. The physical properties of the ink itself affect the pressure fluctuations and movement of the ink in the pressure chambers 212 and nozzles 211, as well as the separation of flying ink from the ink column protruding from the nozzles 211. These may also be affected by interactions between the ink and the pressure chambers 212 and nozzles 211. There are many parameters that can affect these, and these parameters combine to produce the above-mentioned effects. Therefore, by including at least one of these parameters, and as many as possible, and considering their combinations in a comprehensive manner, more accurate estimation results of the ink physical properties, ejection characteristics, and print quality can be obtained.

[0119] The ejection characteristics may also include at least one of the ejection speed, the amount, size, and shape of the ejected ink droplets, the accuracy of droplet landing position (straightness), the number of satellites, and the variation and stability thereof. These may be influenced by not only the ink physical properties but also the interactions between the ink and the pressure chambers 212, the nozzles 211, and the piezoelectric elements 213. There are many parameters that can affect these, and these combine to produce the above-mentioned effects. Therefore, by including at least one of these many parameters, and as many as possible, and considering their combinations both vertically and horizontally, more accurate estimation results of the ink physical properties, ejection characteristics, and print quality can be obtained.

[0120] The waveform information may also include at least one of the following: voltage amplitude, rise time, fall time, amount of overshoot and undershoot relative to the set voltage of the drive signal, distortion rate, amount of harmonics, frequency of harmonics, shape of the frequency spectrum, and numerical data or image data representing the waveform of the electrical signal. As described above, waveform information contains a large amount of information, and quantitative estimation requires quantitative evaluation of this information. For this reason, including at least one of these parameters, or as many as possible, allows for more accurate estimation of the ink properties, ejection characteristics, and print quality.

[0121] Furthermore, the generation and estimation of the database 122 may be performed based on waveform information acquired by outputting a drive signal having the same frequency components from the drive unit 23 to the piezoelectric element 213. If the ink characteristics are similar or the same, more accurate estimation results of the ink properties, ejection characteristics, and print quality can be obtained by adjusting other conditions related to ink ejection.

[0122] The database 122 may also include waveform information when the driver 23 outputs drive signals having different frequency components to the piezoelectric elements 213. The above description is based on the premise that the drive signal has a frequency component related to ink ejection, but depending on the physical properties of the ink, more distinctive waveform information may be obtained in response to other frequency components. Therefore, in this evaluation method, by defining a drive waveform having a plurality of different frequency components and acquiring a corresponding plurality of pieces of waveform information, more accurate estimation results of the ink physical properties, ejection characteristics, and print quality can be obtained.

[0123] The drive signal may also include a drive signal with a lower frequency component than the drive signal output by the drive unit 23 to eject ink. The drive waveform for ejecting ink is determined according to characteristic frequencies such as the resonant frequency of pressure vibration in the ink pressure chamber and the vibration of the liquid surface in the nozzle. Therefore, voltage application using a drive waveform with lower frequency components makes it difficult for ink droplets to be separated and ejected. However, by using a drive waveform with such low frequency components, the effects of characteristics different from the response of the vibration system are more likely to be clearly reflected in the waveform information. Therefore, this evaluation method can obtain estimation results that more comprehensively reflect the ink properties.

[0124] Furthermore, the program 121 of this embodiment causes the computer to function as the following means: (1) an acquisition means for acquiring waveform information of an electrical signal from the inkjet recording apparatus 1; and (2) an estimation means for estimating at least one of the ink physical properties, ejection characteristics, and print quality based on the waveform information. By installing such a program in a computer and having it executed by the control unit 11 or the like, it is possible to obtain accurate estimation results of the ink physical properties, ejection characteristics, and print quality using the inkjet recording apparatus 1, even without a dedicated information processing device.

[0125] The evaluation device of this embodiment also includes a control unit 11. The control unit 11 serves as an information acquisition unit and acquires waveform information of an electrical signal from the inkjet recording device 1. The control unit 11 serves as an estimation unit and estimates at least one of ink physical properties, ejection characteristics, and print quality based on the waveform information. With this evaluation device, more appropriate information about the ink ejected from the inkjet recording device 1 can be easily acquired.

[0126] The inkjet recording device 1, which is a liquid material ejection device of this embodiment, includes the following components: (1) a drive unit 23 that outputs a first drive signal; (2) an ink supply unit 25 that supplies a known ink; (3) a recording element 21 having a piezoelectric element 213, which ejects the known ink by displacing the piezoelectric element 213 in response to the first drive signal output from the drive unit 23, thereby applying the known ink to a medium and performing printing; (4) a waveform information acquisition unit 24 that acquires first waveform information of a first electrical signal detected at the input side of the drive signal to the piezoelectric element 213; and (5) a control unit 11 that functions as a database generation unit that associates the first waveform information with at least one of the physical properties, ejection characteristics, and print quality of the known ink, and generates a database for making estimations related to an ink to be evaluated that is different from a known liquid material. The inkjet recording device 1 can obtain waveform information related to ink ejection operations that are suited to actual conditions by using the drive waveforms, recording element 21, ink supply unit 25, etc. that are actually used for printing operations within the device itself. Therefore, according to the inkjet recording apparatus 1, it is possible to easily obtain more accurate estimation results relating to the characteristics of the ink ejected by the apparatus itself, the ejection characteristics from the nozzles, and the print quality of the ink.

[0127] The control unit 11 of the inkjet recording apparatus 1 also functions as an operation control unit, causing the ink supply unit 25 to supply the liquid material to be evaluated. The control unit 11 functions as an operation control unit, causing the drive unit 23 to output a second drive signal, causing the recording element 21 to print using the ink to be evaluated. The control unit 11 functions as an operation control unit, causing the waveform information acquisition unit 24 to acquire second waveform information of a second electrical signal detected at the input side of the second drive signal to the piezoelectric element 213. The control unit 11 functions as an estimation unit, using a database 122 based on the acquired second waveform information, to determine at least one of the ink properties, ejection characteristics, and print quality to be evaluated as an estimation result for the ink to be evaluated. The control unit 11 functions as an adjustment unit, adjusting the second electrical signal based on a comparison between the estimation result and a set target. The inkjet recording apparatus 1 can use the generated database 122 to determine a drive signal that will achieve desired ejection characteristics or print quality. Therefore, the inkjet recording apparatus 1 can easily achieve more desired ejection characteristics or output a printed matter with desired print quality.

[0128] Furthermore, there may be a plurality of recording elements 21 each having a piezoelectric element 213. The driving unit 23 may be capable of outputting a plurality of different first drive signals in parallel to the plurality of piezoelectric elements 213. The ink supply unit 25 may be capable of supplying different known inks to the plurality of recording elements 21. The waveform information acquiring unit 24 may acquire a plurality of pieces of first waveform information of a plurality of electrical signals detected at the input side of the first drive signals to the plurality of piezoelectric elements 213 in response to the plurality of first drive signals, respectively. The control unit 11 may function as a database generating unit, and may associate the plurality of pieces of first waveform information with at least one of the physical properties, ejection characteristics, and print quality of known inks, to generate a database 122 for making estimations related to an ink to be evaluated that is different from the known ink. When the database 122 is generated for a plurality of ink types, data related to a plurality of different ink types may be obtained in parallel by the plurality of recording elements 21 in this manner. This allows a large number of types of data to be stored in the database 122 efficiently and in a short period of time.

[0129] Furthermore, the plurality of first drive signals may have different frequencies, which makes it possible to accurately estimate ink properties, ejection characteristics, or print quality in a more diverse range of evaluation items, such as the stability of the ink on the medium.

[0130] Alternatively, the control unit 11 of the inkjet recording apparatus 1 functions as an operation control unit, causing the ink supply unit 25 to supply the liquid material to be evaluated. The control unit 11 functions as an operation control unit, causing the drive unit 23 to output one of a plurality of first drive signals as a second drive signal, causing the recording element 21 to print using the ink to be evaluated. The control unit 11 functions as an operation control unit, causing the waveform information acquisition unit 24 to acquire second waveform information of a second electrical signal detected at the input side of the second drive signal to the piezoelectric element 213. The control unit 11 functions as an estimation unit, using the database 122 based on the second waveform information, to determine at least one of the physical properties, ejection characteristics, and print quality of the ink to be evaluated as the result of estimation related to the ink to be evaluated. The control unit 11 functions as an adjustment unit, adjusting the second drive signal based on a comparison between the estimation result and a set target. In this way, the inkjet recording apparatus 1 can determine a drive waveform that achieves desired ejection characteristics or print quality. Depending on the physical properties of the ink, adjusting the frequency component of the drive waveform can also enable more stable and appropriate ink ejection.

[0131] The present invention is not limited to the above-described embodiment, and various modifications are possible. For example, although the database generation and estimation described above includes at least waveform information related to the electrical signal during the voltage application operation of the drive waveform related to ink ejection, the database is not limited to this. The database may be generated and estimated based on at least one of the electrical signals during or after the voltage application operation of the above-described or other various drive waveforms.

[0132] In the above, the analysis process is performed at the end of the database generation process, but the data storage process in the database and the analysis process may be separated and performed separately. In other words, the analysis process does not have to be performed every time a data storage process is performed.

[0133] Furthermore, the database 122 may be separated from the inkjet recording apparatus 1 except when data is registered, analyzed, or estimated. In other words, the database 122 may be configured as a separate entity from the inkjet recording apparatus 1 and used only temporarily. In this case, if the analysis results are stored outside the database 122, the database 122 may be separated from the inkjet recording apparatus 1 even during estimation. Furthermore, a method may be adopted that allows analysis to be performed in real time without generating the database 122 in the first place.

[0134] In the above description, printing is performed using the ink to be evaluated in the ink property estimation process or the ejection characteristics estimation process, but this is not limited to this. The ink to be evaluated may be discharged onto a cover member or the like, and printing may not be performed on the medium.

[0135] 4, the ejection characteristics estimation process shown in Fig. 5, and the print quality estimation process shown in Fig. 6 may be performed in accordance with the intended use, but is not limited to this. A plurality of processes may be performed in parallel or sequentially.

[0136] The electrical signal may also be obtained from an output electrode (terminal) of the driving section 23 to the piezoelectric element 213.

[0137] Furthermore, in the fourth embodiment, the number of piezoelectric elements 213b that perform the output operation of the subtraction-side second signal 232 does not have to be one. The average of the electrical signals of the signal paths connected to the plurality of piezoelectric elements 213b may be used as the electrical signal to be subtracted.

[0138] Furthermore, the frequency components of the drive signal output by the drive unit 23 may be different from the beginning between the database generation and the estimation.

[0139] Furthermore, while the above description has been given of obtaining waveform information corresponding to drive waveforms with different frequency components, the number of such drive signals is not limited to two. It may be three or more. Furthermore, the multiple frequency components may be multiplexed on the frequency axis of the drive waveform, or on the time axis, i.e., different frequency components may be output sequentially in time series. Furthermore, the lowest frequency component among these may have a higher frequency than the lowest frequency component in the drive signal related to ink ejection.

[0140] Furthermore, in the above description, the electromechanical conversion element is the piezoelectric element 213, but it is sufficient if it can convert between an electrical signal and a mechanical displacement.

[0141] In the above description, the storage unit 12, which is composed of a nonvolatile memory such as an HDD or flash memory, has been used as an example of a computer-readable medium for storing the program 121 related to database generation and estimation control of the present invention, but is not limited to this. Other computer-readable media that can be used include other nonvolatile memories such as MRAM and portable recording media such as CD-ROMs and DVD discs. Furthermore, a carrier wave can also be used as a medium for providing program data related to the present invention via a communication line. In addition, the specific configurations, contents and procedures of the processing operations, etc. shown in the above embodiments can be modified as appropriate without departing from the spirit of the present invention. The scope of the present invention includes the scope of the invention described in the claims and its equivalents.

[0142] [Basic concept and research and development history of the present invention] 1. Trends and requirements for manufacturing digital transformation 1-1 Benefits of applying data-driven methods to manufacturing Traditional "manufacturing" has generally been accomplished by skilled workers, such as experts and craftsmen, who use their own experience and insight as a foundation to address problems using a deductive approach, or "forward direction." In other words, traditional methods are highly dependent on the sense of the leader, and are not necessarily efficient, as they often require repeated trial and error. On the other hand, there is also an approach known as "abduction (hypothesis testing)" for research and development and technological development, which is recommended as the most effective development method, especially in academic fields.

[0143] Data-driven methods differ from both "deduction" and "abduction" in that they are approaches that inductively gain insights from acquired data and derive solutions; they are also known as "backward" or "inverse problem solving." The most familiar and large-scale successful example of inverse problem solving is e-commerce, which utilizes big data. The so-called "production DX (digital transformation)" seeks to apply concepts similar to e-commerce to manufacturing. This approach has attracted attention in the industrial world as a groundbreaking method. However, in reality, it remains little more than a hype, and traditional methods remain mainstream in the field, leading to derision such as the "wall of legacy."

[0144] At the research level, the NEDO project "Ultra-Advanced Materials Ultra-Rapid Development Platform Technology Project" (abbreviated as "Ultra-Ultra Project") was implemented over a six-year period beginning in fiscal 2016. In this Ultra-Ultra Project, 18 participating companies and the National Institute of Advanced Industrial Science and Technology (AIST) each focused on different materials and worked to demonstrate chemical manufacturing digital transformation in Japan. This project is a unique attempt to streamline development time and the number of prototypes by utilizing a data-driven approach that combines measurement, calculation, and process. As a result, it was estimated that development time could be reduced to an average of 1 / 23 of the conventional timeframe across all themes. This fact makes clear that the importance and impact of data-driven manufacturing in manufacturing is even greater than that of e-commerce, and it is certain that this approach will dominate the industrial world in the near future.

[0145] 1-2 Significance and problems of data-driven methods in complex manufacturing Manufacturing in the industrial sector encompasses buildings, houses, ships, automobiles, robots, printing equipment, video equipment, cameras, and even the components that make up these. Products are made by designing and processing these components. In manufacturing, where machines and structures are the final products, factory automation, such as measuring the size of parts and assembling them, has been implemented on a large scale from an early stage. For this reason, it is relatively easy to generate data during the manufacturing process and implement data-driven production, which is already being implemented around the world.

[0146] On the other hand, when manufactured products are composites made from chemical substances or materials, such as those in the ultra-ultra-project mentioned above, their components include micron-sized fillers and fibers, submicron- to tens-nanometer-sized polymers, and even substances ranging in size from Å to several nm, such as low-molecular-weight functional substances and low-molecular-weight additives. These components form complex systems that combine through minute bonds and interactions in the Å to nm range, such as hydrogen bonds and van der Waals bonds. Furthermore, the characteristics and properties of these components change in various ways depending on the actual manufacturing conditions and usage environment. This makes it extremely difficult to identify the determinants of desired characteristics and performance.

[0147] In the past, the approach to solving such problems was for experts in each field to deductively understand phenomena by performing scientific measurements and then combine these multiple understandings of phenomena to construct knowledge in a forward-looking manner. However, in the field of bioscience, which has been booming in recent years, things more complex than chemical substances are beginning to appear, such as synthetic biology, which uses biological substances such as microorganisms and enzymes. Therefore, it is no longer possible to obtain the desired results using deductive methods.

[0148] The field of chemistry is inherently less logical than the technical fields of physics, electricity, and mechanics. Furthermore, in the pharmaceutical field, biopharmaceuticals such as antibody drugs and nucleic acid drugs such as mRNA are increasingly replacing small molecule compounds. In other words, a large number of complex substances are emerging that cannot be fully evaluated or measured using conventional definitive testing or rigorous analysis. As the objects we handle transition from simple to complex systems, we are reaching an inflection point where deductive methods and the conventional measuring devices based on them can no longer be used. As a result, we are faced with the urgent need to discover new evaluation technologies and methodologies. This is the greatest motivation for applying data-driven methods to the fields of chemistry and pharmaceuticals.

[0149] In order to transition to a data-driven approach to complex manufacturing, it is essential to quickly and easily digitize the phenomena occurring in complex systems without human intervention. However, in complex systems, it is often unclear which elements of an object are responsible for which functions, and their relationships can change due to the influence of the surrounding environment. For this reason, no matter how much data is collected, it may be "poor quality" data that does not contain correct answers. Even if machine learning is performed based on such data, there is a risk that an appropriate solution will not be found. This is hindering its widespread adoption in industry.

[0150] In other words, when applying data-driven methods to complex systems, it is extremely important to capture the elements and their combinations that are the source of functionality and to obtain "high-quality" data that matches the actual conditions of use.

[0151] 1-3 Data requirements for data-driven methods Practicing machine learning and other techniques requires large amounts of data. For example, deep learning uses models with high degrees of freedom to explain complex phenomena. It is said that in order to make statistically meaningful inferences using these models, data at least 10 times the number of degrees of freedom of the model is required. However, no matter how large the amount of data, unless it is "high-quality" and includes correct answer data, it is not easy to improve the predictive accuracy of machine learning and other techniques.

[0152] Furthermore, machine learning presupposes a finite-dimensional linear space as input and output. In other words, machine learning can map ultra-multidimensional data into that space, but to take advantage of this feature, as much diverse data as possible is required. As mentioned above, it is not easy to predict and define in advance which data will be useful for the target variable, especially in complex manufacturing, and there is a possibility that unexpected explanatory variables may be influencing it. For this reason, collecting multivariate data, even data whose relationship with the target variable is not obvious, as a candidate explanatory variable, is important for digital transformation in complex manufacturing.

[0153] However, collecting a large amount of diverse data generally takes a considerable amount of time and money, so high-throughput data generation that can collect data quickly and inexpensively is an essential requirement for realizing manufacturing DX.

[0154] 1-4 Real-time measurement as a data acquisition method and its requirements In the super-super project mentioned above, the goal of reducing development time to 1 / 20 of the previous timeframe was achieved for all themes. The main factor behind this was the use of simulation to generate massive amounts of data. In other words, if extremely large amounts of data are generated through simulation, there is a high possibility that the various data-related requirements mentioned above can ultimately be resolved. However, simulation is not a panacea. Simulation has problems such as the need for physical models, the fact that model selection is subjective, and the limited versatility of the models necessitates the selection of adaptation laws appropriate to the target phenomenon and the adjustment of simulation parameters. In other words, using simulation requires know-how and time, making it difficult to popularize.

[0155] One approach to solving this problem is to measure real objects while setting multiple conditions and acquiring large amounts of multi-type, multivariate data. If large amounts of data can be acquired through measurement, the quality of the data can be guaranteed to a certain extent. However, when the object is a liquid or fluid, there are issues such as the fact that liquids and fluids are more difficult to handle than solids and are more susceptible to environmental influences such as temperature. This can make it difficult to acquire "high-quality" data quickly, easily, and inexpensively, which is a bottleneck in the transition to data-driven development.

[0156] Automating data acquisition using dedicated devices and equipment makes it possible to quickly and easily obtain large amounts of data. However, because such devices and equipment are dedicated, they tend to increase costs. If hardware factors such as measuring equipment become rate-limiting when acquiring large amounts of data, it can no longer be called DX. Furthermore, with liquids, which are easily affected by their environment, there is a tendency for a gap to form between the dedicated measurement environment and the environment in which the object is actually used. As a result, there is a risk that "high-quality" data cannot be obtained.

[0157] Manufacturing is thus at a major turning point with the SDGs, the super-smart society, and Web 3.0. In this kind of manufacturing, measurement methods that can quickly, easily, and inexpensively obtain the high-quality data necessary for data-driven methods from liquid or fluid-like objects, such as chemical substances and bioproducts, are essential. And the methodology that embodies this is thought to be an invention.

[0158] 2. The role of data-driven methods in inkjet printer ink development 2-1 Quality requirements for inkjet ink Inkjet printers are characterized by their ability to produce images easily and inexpensively, and are used in a variety of fields, including photography, various printing methods, marking, and special printing such as color filters. The inkjet method uses digital printing without the use of plates. Therefore, the inkjet method is particularly suitable for applications that require the production of a variety of images in small quantities, and can be said to be an essential technology for realizing a super-smart society in which "what is needed, to the people who need it, at the time they need it, and only what is needed."

[0159] There are various types of inks used in inkjet printing (hereinafter referred to as inkjet inks, or simply inks). For example, there are water-based inks made of water and a small amount of organic solvent, non-water-based inks containing organic solvents but substantially no water, hot-melt inks that are printed by heating and melting inks that are solid at room temperature, and actinic radiation-curable inks that are cured by irradiation with actinic radiation after printing. Alternatively, inks can be classified into types based on the color-developing principle, etc. These various inks are used according to the application and image medium.

[0160] Inkjet printers employ a mechanism for ejecting ink droplets to form images, and therefore inks are required to meet extremely high and diverse quality standards. Accordingly, various requirements must be met to achieve these quality standards. Typical requirements include image quality that exhibits color tone and clarity on the image medium, image durability (e.g., lightfastness and waterfastness) to prevent deterioration or breakage of the ink coating on the image during storage or post-processing of the printed matter, and appropriate liquid properties for ejecting ink droplets from the ink ejection mechanism within the image-forming device. However, these requirements vary depending on the application. Furthermore, these requirements often have a trade-off relationship, creating a complex situation in which improving one requirement can result in a deterioration in the quality of other requirements.

[0161] Similarly, in terms of the chemical composition of ink, a complex combination of ingredients is required to achieve specific qualities, but these ingredients interact with each other, making it difficult to maintain the appropriate balance. For this reason, changes to the chemical composition can cause unexpected side effects, and it is not easy for developers to understand these phenomena and create a comprehensive design.

[0162] Furthermore, ink is a liquid and is easily affected by the surrounding environment, such as temperature, and the requirements for its properties, quality, composition, and ingredients can vary widely.

[0163] As described above, inkjet inks are complex substances that can only be made by properly balancing the numerous requirements, properties, and compositions that are in a trade-off relationship with one another. Therefore, their development and manufacturing require advanced design methods that can resolve these issues in a comprehensive manner.

[0164] 2-2 Challenges in ink development (1) ~Cost issues, inconsistencies in environmental conditions~ As mentioned above, inkjet printers must satisfy various requirements, taking into account factors such as ink ejection performance and post-print storage stability. Among these, the physical properties of the ink are particularly important, including viscosity (viscosity), elasticity, viscoelasticity, density, specific gravity, solids concentration, surface tension, thixotropy, gas content, dielectric constant, charge rate, refractive index, zeta potential, dispersed particle size, and particle size distribution. Viscosity, among other properties, is considered a particularly important parameter in the ink ejection mechanism design, and various dedicated devices and measuring instruments for measuring viscosity have been put into practical use. For example, Japanese Patent Application Laid-Open No. 2009-174930 (Prior Document 1) and Japanese Patent Application Laid-Open No. 2005-61871 (Prior Document 2) disclose a method for measuring ink viscosity by utilizing the phenomenon in which the resonant frequency and electrical impedance of a piezoelectric element change when ink is dropped onto the element.

[0165] However, these devices are generally expensive and require a separate measurement environment. Therefore, measurement work becomes a preliminary or intermediate step in the ink development and manufacturing process, which can result in unnecessary labor and costs. Furthermore, even if these dedicated measurement devices use the same piezoelectric elements as inkjet printers, their structure and operating conditions are completely different from those of actual inkjet printers. Furthermore, these measurement devices are inevitably operated in environments that differ from those of actual inkjet printers. As a result, the values ​​obtained by these measurement devices do not necessarily fully describe the actual ink physical properties in inkjet printers.

[0166] 2-3 Challenges in Ink Development (2) ~Complex Systems, Trade-Offs, and Substitute Properties~ When developing inkjet inks, which are complex systems, it is important to resolve trade-offs between quality requirements and predict the complex phenomena of chemical composition. Applying data-driven methods to development is expected to solve this problem, but in this case, appropriate evaluation criteria and measurement methods must be established. However, even at this stage, trade-offs and appropriate balances between many requirements and elements are required, and optimizing them to achieve better quality and characteristics is extremely difficult.

[0167] The inventors concluded that the trade-offs between these elements and requirements were due to the fact that all decisions were left to human judgment. In other words, the ink-related physical properties that have been measured and considered appropriate characteristics are not necessarily essential, but merely a limited perspective on the complex composition of ink. Human deductive thinking, which can only visualize things in three dimensions at best, makes it extremely difficult to find an optimal solution from conflicting data. Despite this, humans have interpreted these physical properties as if they were the essence of the ink and attempted to find a solution from that perspective. The inventors concluded that this poses a fundamental problem. If we could find an optimal solution based solely on data, without relying on human understanding—that is, if we adopted the results of big data analysis in science—humans should be able to arrive at a solution without experiencing any trade-off dilemmas. This is what digital transformation in science and manufacturing is all about.

[0168] Furthermore, there is information and data that has not received much attention until now because it has been deemed meaningless or incomprehensible to humans. Based on the above philosophy, if this information is actually the product of interactions and influences between inkjet printer elements and factors, it can be considered highly useful data for machine learning.

[0169] Traditionally, inkjet inks have been developed by measuring their main physical properties, such as viscosity, offline, independently of the inkjet printer, and linking the physical properties obtained through this measurement to print image quality. However, as mentioned above, in complex systems, the physical properties themselves are no longer essential, but merely substitute characteristics introduced by humans based on the expectation that they will have an impact. The current state of development is one in which traditional deductive methods are used without understanding this, and most engineers are unaware of the major problem that lies here.

[0170] The essence of this invention is to break through the preconceived notions mentioned above through data-driven methods and to provide a method for optimizing elements and factors in complex systems that at first glance appear to be trade-offs.

[0171] 2-4 What data is required for ink development digital transformation? Data acquired under conditions and environments that differ from the specifications, structure, and operating environment of an actual printer does not take into account important factors such as contact and convection between the ink and the inkjet head surface, and other interactions between the ink and the inkjet head. Therefore, this data cannot be considered "high quality" for use in data-driven methods, such as machine learning. In contrast, data acquired under the original usage patterns and conditions, in which ink is filled into the inkjet head and ejected, is essentially "high quality." Therefore, it can be said that there is great significance and meaning in deriving solutions using machine learning and other methods based on this high-quality data.

[0172] Considering the prior art from this perspective, Japanese Patent Application Laid-Open No. 2007-030344 (Prior Art. 3) and Japanese Patent Application Laid-Open No. 2012-006310 (Prior Art. 4) disclose inventions that use the electrical characteristics of an inkjet head to obtain measurement data and predict or use the ink's physical properties (viscosity) as a substitute. These methods obtain data that serves as the basis for developing deductive thinking through measurement, with the aim of helping humans understand the mechanisms and principles of ink ejection. In other words, the prior art attempts to obtain data on substitute characteristics by using an inkjet head as the measurement environment. Therefore, these prior arts should be clearly distinguished from the present invention, which treats various information obtained from a complex system as a complex system and attempts to mechanically derive a solution.

[0173] The state of ink ejected from an inkjet head and the printed image formed with the ejected ink are the result of the interaction between various ink and inkjet head characteristics. From this perspective, the various measurements, such as viscosity, referred to as substitute characteristics, can also be considered data obtained as a result of some action being taken on ink as it passes through the inkjet head. In other words, the various pieces of information obtained from the inkjet head are linked to various characteristics related to ink ejection and print image formation, including substitute characteristics, even if they are beyond human comprehension. In this sense, all of this information can be considered data that can be analyzed using techniques such as machine learning. Furthermore, setting values ​​such as the specifications of the electrical signal applied to the inkjet head (waveform, frequency, phase, etc.) can be parameterized. Waveform information obtained in response to ink ejection operations, as well as differences and feature values ​​relative to changes from the original electrical signal, can also be obtained as data. Even if these are too complex for humans to understand, AI techniques such as machine learning can extract useful data and provide more accurate solutions. This, too, is a different concept from conventional ink evaluation methods and inkjet head utilization methods, and offers an advantageous feature.

[0174] 3. Liquid Dispenser as a Data Generation Environment for Data-Driven Development 3-1 Background and thinking behind the present invention The present invention has been made in view of the above problems and circumstances. The background to the invention will be explained below.

[0175] As described above, the piezoelectric element used in an inkjet head converts an applied electrical signal into a mechanical displacement, thereby achieving the function of pushing out supplied ink. Conversely, the piezoelectric element has the reversible property of converting a mechanical displacement into an electrical signal when applied. The dedicated devices and apparatuses shown in Prior Art Documents 1 and 2 also utilize the effect of the piezoelectric element to measure the viscosity characteristics of ink, etc.

[0176] On the other hand, Prior Art Documents 3 and 4 use the inkjet head as a sensor. That is, they show a method of detecting residual vibrations that occur in a piezoelectric element after ink is ejected and estimating the viscosity of the ink from this residual vibration. However, in an inkjet head, the residual vibrations are merely a phenomenon that occurs after ink is ejected, and do not directly represent the characteristics of the ink at the time of ejection. As such, the inventions described in the prior art documents measure or predict only viscosity, which is a substitute characteristic. Furthermore, these inventions predict viscosity indirectly from data after ink ejection, rather than directly acquiring data during ejection, which likely results in a deterioration in the quality of the data. Furthermore, for complex liquid samples, it is difficult to acquire large amounts of data in a short period of time due to the laborious process of determining and setting the conditions, making these methods completely insufficient for acquiring data to be analyzed using AI technologies such as machine learning.

[0177] On the other hand, if AI technology is used, all data will be processed by computers, so the data itself does not necessarily need to be understandable to humans. Data-driven methods shift the decision-making focus from humans to machines, making it possible to use all kinds of data, including ultra-multidimensional data that was previously difficult to handle. In other words, data-driven methods use a wide variety of multi-variate data obtained from complex systems to estimate their properties, phenomena, structure, composition, and more. This is expected to remove constraints on research and development and manufacturing and dramatically improve their efficiency.

[0178] Therefore, the inventors focused on the trend of data-driven methods and their data requirements and came up with a method for quickly and easily obtaining direct and "high-quality" data on ink characteristics and the like from a device that is actually in the ink ejection process.

[0179] That is, in this method, a drive signal of a predetermined waveform is applied to an inkjet head to drive it, and various information regarding the waveform of the electrical signal during the ejection operation is acquired. When ink is supplied and ejected, the mechanical displacement of the piezoelectric element is affected by various properties of the ink, such as the viscosity of the ink, causing fluctuations in the electrical internal state (e.g., impedance). As a result, the voltage division ratio between the piezoelectric element and a drive circuit, such as a driver amplifier connected to the inkjet head, fluctuates, causing the amplitude of the electrical signal at the input terminal of the piezoelectric element to change. Furthermore, if delays occur in the displacement operation of the piezoelectric element due to the viscosity or elasticity of the ink, the electrical signal waveform will change in a more complex manner, accompanied by capacitive and inductive phase shifts.

[0180] As described above, the electrical signal waveform applied to the piezoelectric element and detected at the input terminal changes and is determined by a complex combination of multiple factors, including the drive signal waveform that was originally applied, the electrical characteristics of the piezoelectric element that change due to factors such as the ink properties and the compatibility between the ink and the piezoelectric element, and the environmental conditions in which these are placed. It is difficult to manually handle such complex, multidimensional, multivariate data. However, the inventors believed that by utilizing AI technology, it would be possible to analyze this data in its entirety, evaluate the formulation and characteristics of the ink actually ejected, as well as the ink ejection characteristics from the nozzle, and ultimately predict the resulting print image quality.

[0181] Furthermore, this data is acquired using the actual inkjet printer itself as the acquisition environment, under conditions that are identical to or similar to the actual usage environment. This makes it easy to ensure consistency between the actual usage environment and the evaluation environment in terms of piezoelectric element specifications, drive conditions, compatibility between ink and piezoelectric element, environmental parameters, etc., and essentially makes it possible to acquire "high-quality" data.

[0182] Furthermore, inkjet printers generally have a large number of nozzles in their inkjet heads. The inventors envisioned that by utilizing this structure as a multi-channel sensor device and simultaneously ejecting multiple inks with different characteristics under different driving and environmental conditions, it would be possible to rapidly acquire large amounts of multidimensional and multivariate data. For example, by dividing a head with a total of 1,024 nozzles into 32 sections and ejecting inks with different pigment types and ink compositions, it would be possible to realize a high-throughput data generation device that can simultaneously acquire information on 32 types of ink compositions.

[0183] In this way, the ink evaluation method according to the present invention can improve the recurrence of data (reliability of the solution) by utilizing the abundant information contained in the waveform information obtained from the inkjet head. Furthermore, considering the essence of the present invention, it is even more effective for ink evaluation to utilize various data related to the ink, inkjet head, or printed image obtained by other different measurement methods or signal generation means in addition to the waveform information from the inkjet head. Another feature is that these measurement modalities can be used in combination.

[0184] The above approach is an idea that only inventors who develop and manufacture their own inkjet heads, which are key devices in inkjet printers, could come up with. In other words, it is something that would never be thought up by someone who purchases inkjet heads and builds equipment.

[0185] 3-2 Social value brought by this invention This invention contributes to the efficiency of ink development and evaluation by acquiring a large amount of various information about inkjet inks using an inkjet head or inkjet printer that ejects the ink itself, thereby obtaining a large amount of "high-quality" data for machine learning and other purposes in a short amount of time, without the need for complex operations or dedicated equipment. Furthermore, this is a revolutionary ink formulation determination method that allows for the rapid development of appropriate inks, even in the face of the challenge that the printed surface (media) changes depending on the customer and industry, which is the fate of ink.

[0186] For many inkjet-printed materials, the performance and commercial value desired by users lie not in the ink's physical properties or composition, but in the quality and characteristics of the resulting printed image. As with the above-mentioned printed media, the present invention can easily accommodate inks with modified pigments, even when the spectral absorption characteristics of the dyes added to the ink are adjusted for a specific application. These features can be applied to the development of a variety of inks, with machine learning being common to both dye-based and pigment-dispersed inks.

[0187] As described above, the present invention makes it possible to quickly and rationally ensure compatibility between inkjet heads and ink. Furthermore, the present invention can also contribute to the development of high-performance, high-function inks that contribute to improving the quality of printed images. From another perspective, it is also possible to statistically weight and identify various waveform information and its feature quantities associated with ink obtained in this way. This has the potential to be applied as a new method for evaluating ink characteristics or ejection characteristics, and ultimately as a new method for evaluating print image quality.

[0188] Furthermore, the inkjet head can be said to be an excellent inspection device that can generate overall feature quantities in the form of electrical signals for any liquid material, including solutions or dispersions, other than printing liquids. In other words, the present invention is not limited to on-demand printing applications, but can also be applied to future data generation devices for machine learning on liquid objects. [Explanation of symbols]

[0189] 1, 1a to 1e Inkjet recording device 11 Control section 12 Storage section 121 Programs 122 databases 13 Communications Department 14 Display section 15 Operation reception section 16 Measurement section 20, 20a to 20e Recording operation section 21, 21a to 21c recording elements 211, 211a to 211c nozzles 212, 212a-212c pressure chamber 213, 213a to 213c Piezoelectric element 22, 22a, 22d, 22e signal generating unit 23, 23a, 23b, 23e Drive unit 24 Waveform information acquisition section 25 Ink supply unit 26 Transfer Channel M Machine Learning Model

Claims

1. 1. An evaluation method for a liquid material discharged by a liquid material discharge device, the evaluation method comprising: a drive unit; and a recording element having an electromechanical conversion element, the recording element displacing the liquid material by outputting a drive signal from the drive unit to displace the electromechanical conversion element, and applying the liquid material onto a medium to perform printing, the method comprising: An evaluation method for estimating at least one of the characteristics, ejection characteristics, and print quality of the liquid material based on waveform information of an electrical signal detected on the input side of the drive signal to the electromechanical conversion element.

2. The evaluation method according to claim 1, further comprising generating a database relating to the correspondence between the waveform information when a known liquid material is ejected and the printing is performed and at least one of the characteristics of the known liquid material, the ejection characteristics, and the print quality, and performing the estimation using the database.

3. the liquid discharge device has a plurality of the recording elements, The evaluation method according to claim 1, further comprising generating a database relating to the correspondence between the waveform information and at least one of the characteristics, ejection characteristics, and print quality of the known liquid material corresponding to each of the plurality of recording elements when printing is performed using the at least two types of known liquid material, each having different characteristics, from the plurality of recording elements, and performing the estimation using the database.

4. Acquire the waveform information when the liquid material to be evaluated is discharged and the printing is performed; estimating at least one of the characteristics, ejection characteristics, and print quality of the liquid material to be evaluated based on the waveform information using the database; The evaluation method according to claim 2 or 3.

5. 2. The evaluation method according to claim 1, wherein the waveform information is acquired from the electrical signal when the driving unit is outputting a driving signal that causes the recording element to eject the liquid material.

6. 2. The evaluation method according to claim 1, wherein the waveform information is obtained from the electrical signal after the driving unit has output a driving signal that causes the recording element to eject the liquid material.

7. The evaluation method according to claim 1 , wherein the estimation is performed using a machine learning model that outputs at least one of the characteristics of the liquid material, the ejection characteristics, and the print quality in response to the input of the waveform information.

8. The evaluation method according to claim 1 , wherein the electrical signal is detected from a path electrically connecting the driving section and the electromechanical conversion element.

9. the liquid discharge device has a plurality of the recording elements, the driving unit can select whether or not to output the driving signal to each of the electromechanical conversion elements included in the plurality of recording elements, The evaluation method described in claim 1, wherein the electrical signal is detected from a path electrically connecting the drive unit and a second electromechanical conversion element possessed by the second recording element when the liquid material ejected in response to the displacement of a first electromechanical conversion element possessed by the first recording element is guided to a second recording element different from the first recording element.

10. the liquid discharge device has a plurality of the recording elements, the driving unit can select whether or not to output the driving signal to each of the electromechanical conversion elements included in the plurality of recording elements, the waveform information includes a difference between the electrical signal detected from a path electrically connecting the driving unit and a third electromechanical transducer element of the third recording element and an electrical signal detected from a path electrically connecting the driving unit and a fourth electromechanical transducer element of the fourth recording element; The evaluation method according to claim 5.

11. 2. The evaluation method according to claim 1, wherein the physical properties of the liquid material include at least one of viscosity, elasticity, density, specific gravity, solid content, surface tension, thixotropy, gas content, dielectric constant, charge rate, refractive index, zeta potential, dispersed particle size, and particle size distribution.

12. 2. The evaluation method according to claim 1, wherein the ejection characteristics include at least one of an ejection speed, an amount, a size and a shape of an ejected ink droplet, an accuracy of a landing position, a number of satellites, and a variation and stability thereof.

13. 2. The evaluation method according to claim 1, wherein the waveform information includes at least one of a voltage amplitude, a rise time, a fall time, an amount of overshoot and an amount of undershoot relative to a set voltage of the drive signal, a distortion factor, an amount of harmonics, a frequency of harmonics, a shape of a frequency spectrum, and numerical data or image data representing the waveform of the electrical signal.

14. The evaluation method according to claim 4 , wherein the database generation and the estimation are performed based on the waveform information obtained by outputting the drive signal having the same frequency component from the drive unit to the electromechanical conversion element.

15. 4. The evaluation method according to claim 2, wherein the database contains the waveform information when the driving section outputs drive signals having different frequency components to the electromechanical transducer element.

16. The evaluation method according to claim 15, wherein the drive signal includes a drive signal having a frequency component lower than that of the drive signal output by the drive unit for ejecting the liquid material.

17. Computer, an acquisition means for acquiring waveform information of an electric signal detected on the input side of the drive signal to the electromechanical conversion element from a liquid discharge device including a drive unit and a recording element having an electromechanical conversion element, the recording element outputting a drive signal from the drive unit to displace the electromechanical conversion element, thereby discharging a liquid material and applying the liquid material onto a medium to perform printing; an estimation means for estimating at least one of the characteristics, ejection characteristics, and print quality of the liquid material based on the waveform information; A program that functions as a

18. an information acquiring unit that acquires waveform information of an electrical signal detected on the input side of the drive signal to the electromechanical conversion element from a liquid discharge device that includes a drive unit and a recording element that has an electromechanical conversion element and outputs a drive signal from the drive unit to displace the electromechanical conversion element to discharge a liquid material and apply the liquid material onto a medium to perform printing; an estimation unit that estimates at least one of the characteristics, ejection characteristics, and print quality of the liquid material based on the waveform information; An evaluation device comprising:

19. a drive unit that outputs a first drive signal; a liquid supply unit that supplies a known liquid; a recording element having an electromechanical conversion element, displacing the electromechanical conversion element in response to the first drive signal output from the drive unit to eject the known liquid material and apply the known liquid material onto a medium to perform printing; an information acquiring unit that acquires first waveform information of a first electrical signal detected at an input side of the drive signal to the electromechanical conversion element; a database generation unit that associates the first waveform information with at least one of the characteristics, ejection characteristics, and print quality of the known liquid material, and generates a database for making an estimation related to a liquid material to be evaluated that is different from the known liquid material; A liquid material ejection device comprising:

20. an operation control unit that causes the liquid material supply unit to supply the liquid material to be evaluated, causes the drive unit to output a second drive signal, causes the recording element to print using the liquid material to be evaluated, and causes the information acquisition unit to acquire second waveform information of a second electrical signal detected at an input side of the second drive signal to the electromechanical conversion element; an estimation unit that determines, based on the second waveform information, at least one of the characteristics, ejection characteristics, and print quality of the liquid material to be evaluated using the database, as an estimation result related to the liquid material to be evaluated; an adjustment unit that adjusts the second drive signal based on a comparison between the result of the estimation and a set target; The liquid material discharge device according to claim 19, comprising:

21. a plurality of recording elements each having the electromechanical conversion element; the drive section is capable of outputting a plurality of different first drive signals in parallel to a plurality of the electromechanical conversion elements, the liquid supply unit is capable of supplying different known liquid materials to the plurality of recording elements; the information acquisition unit acquires a plurality of first waveform information of a plurality of electrical signals detected at input sides of the first drive signals to the plurality of electromechanical conversion elements in response to the plurality of first drive signals, respectively; the database generation unit associates the plurality of pieces of first waveform information with at least one of the characteristics, ejection characteristics, and print quality of the known liquid material, and generates a database for making an estimation related to a liquid material to be evaluated that is different from the known liquid material. The liquid material discharge device according to claim 19.

22. the plurality of first drive signals have different frequencies from each other; The liquid material discharge device according to claim 21.

23. an operation control unit that causes the liquid material supply unit to supply a liquid material to be evaluated, causes the drive unit to output any one of the plurality of first drive signals as a second drive signal, causes the recording element to print using the liquid material to be evaluated, and causes the information acquisition unit to acquire second waveform information of a second electrical signal detected at an input side of the second drive signal to the electromechanical conversion element; an estimation unit that determines, based on the second waveform information, at least one of the characteristics, ejection characteristics, and print quality of the liquid material to be evaluated using the database, as an estimation result related to the liquid material to be evaluated; an adjustment unit that adjusts the second drive signal based on a comparison between the result of the estimation and a set target; The liquid material discharge device according to claim 22, comprising:

Images