Process state monitoring system for new energy battery manufacturing
By constructing a process status monitoring system, the liquid column status can be identified and adjusted in real time, solving the problem of abnormal liquid column in the manufacturing of new energy batteries and achieving stability and consistency in the liquid injection process.
Patent Information
- Application Number
- CN202511204332.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-08-26
- Publication Date
- 2025-12-05
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
In the manufacturing process of new energy batteries, problems such as intermittent or interrupted flow and air bubble entrainment are prone to occur during the liquid injection process, resulting in insufficient liquid injection or unwetting of air bubbles. Moreover, the existing control methods lack the ability to monitor the internal structure of the liquid column in real time, making it impossible to make timely adjustments and causing potential quality hazards.
A process status monitoring system is constructed, including a disturbance excitation module, a response acquisition module, a status identification module, and a control command generation module. The system excites the liquid column with disturbance signals, acquires response signals in real time, identifies abnormal liquid column status, and generates corrective control commands to achieve closed-loop process monitoring.
It enhances the ability to actively sense the state of the liquid column, improves the accuracy and response speed of anomaly identification, reduces process errors, and ensures the stability and consistency of the injection process.
Smart Images
Figure CN121076183A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of new energy battery production, and particularly relates to a process state monitoring system for new energy battery manufacturing. BACKGROUND
[0002] In the manufacturing process of new energy batteries, electrolyte injection is an important process link for ensuring the capacity, cycle life and safety performance of the battery cell. The process is usually carried out in a vacuum or dry environment, and a quantitative electrolyte is injected into the packaged battery cell structure through an injection needle. In actual production lines, the flow state of the liquid column in the injection channel is disturbed by multiple factors, including liquid viscosity fluctuations, channel blockage, pressure disturbances or unstable equipment responses; as a result, the following problems are prone to occur during the injection process:
[0003] The liquid column is intermittent or the flow is interrupted, resulting in insufficient injection volume;
[0004] Air bubbles are entrained into the inter-electrode gap, causing local non-wetting;
[0005] Injection abnormalities are delayed to be discovered, and the packaged battery cell cannot be corrected, resulting in the entire battery cell being scrapped.
[0006] The current injection control method is mainly based on timing, pressure or quantity strategy, which can ensure the completion of surface injection, but lacks real-time monitoring capability for the internal structure state of the liquid column. Especially during the injection process, whether the liquid column has a non-continuous region (such as an air-entraining section) cannot be directly identified by traditional liquid level or flow sensors, which may cause quality problems.
[0007] In addition, the current process state recognition and execution action are disconnected, and there is a lack of a systematic structure to effectively connect the state recognition module and the injection control device. This leads to the fact that even if an abnormality occurs, the injection action cannot be adjusted in the process. SUMMARY
[0008] The purpose of the present application is to provide a process state monitoring system for new energy battery manufacturing, which should have the following capabilities: active excitation and response collection capability for the liquid column state, state recognition capability based on response behavior, and process control capability for issuing correction instructions to the injection execution module, thereby constructing a process monitoring closed loop from abnormality recognition to action response.
[0009] The above technical purpose of the present application is achieved by the following technical solution:
[0010] A process state monitoring system for new energy battery manufacturing, comprising:
[0011] A disturbance excitation module is arranged outside the structure of the electrolyte injection path and is used to apply a controllable disturbance signal to the liquid column during the injection process;
[0012] A response acquisition module, corresponding to the disturbance excitation module, is configured to acquire a response signal of the liquid column under the action of the disturbance;
[0013] A state identification module, connected to the response acquisition module, is configured to identify a state of the liquid column based on the response signal and determine whether there is a non-continuous region of the liquid column or a bubble entrainment behavior;
[0014] A control instruction generation module, connected to the state identification module, is configured to generate a correction control instruction when the state of the liquid column is determined to be abnormal;
[0015] A liquid injection execution module, connected to the control instruction generation module, is configured to receive the correction control instruction and perform a liquid injection pause, air exhaust, or liquid injection rate adjustment operation.
[0016] Further provided are:
[0017] The disturbance excitation module includes:
[0018] A beat identification logic module is configured to receive a control cycle signal from the liquid injection execution module and identify an intermittent phase in a liquid injection cycle based on the signal;
[0019] A disturbance pulse generation path, connected to the beat identification logic module, is configured to receive a trigger signal in the intermittent phase and control a disturbance pressure source or an array of liquid injection control valves to generate a disturbance pulse signal with an amplitude lower than a main pressure of the liquid injection;
[0020] A timing control strategy unit, connected to the beat identification logic module and the disturbance pulse generation path, is configured to set a trigger time point and an action duration of the disturbance pulse signal in the intermittent phase, so that the disturbance pulse signal is applied to the liquid column in a center-symmetric manner through the liquid injection channel and is used to excite the response signal of the liquid column under the action of the disturbance.
[0021] Further provided are:
[0022] The disturbance pulse generation path determines the amplitude and period of the disturbance pulse signal based on a disturbance control model, which is constructed according to physical properties of the liquid column or structural parameters of the liquid injection path, so that the generated disturbance pulse signal has dynamic adaptive characteristics sufficient to excite the response signal of the liquid column and does not interfere with the liquid injection process.
[0023] Further provided are:
[0024] The disturbance control model includes an amplitude function and a period function,
[0025] The amplitude function A adopts an exponential decay form and is constructed based on the dynamic viscosity μ, the density ρ, and the length L of the liquid injection path as follows:
[0026]
[0027] The periodic function T adopts a segmented response form, is adjusted based on the path length L and the natural frequency interval of the liquid column, and an expression is as follows:
[0028]
[0029] Wherein,
[0030] A0 is a reference disturbance amplitude;
[0031] λ is an empirical attenuation coefficient;
[0032] T1, T2, L0 are empirical setting constants;
[0033] e is the base of the natural logarithm;
[0034] log is a natural logarithm function.
[0035] Further settings are as follows:
[0036] The response acquisition module comprises at least one non-contact vibration sensor, for acquiring the vibration response signal generated after the disturbance pulse signal acts on the liquid column; the response acquisition module realizes synchronous acquisition by receiving a trigger signal from the disturbance excitation module, and performs preset signal denoising and amplitude normalization processing on the acquired vibration response signal.
[0037] Further settings are as follows:
[0038] The state recognition module comprises a frequency stability analysis unit and an attenuation feature extraction unit,
[0039] The frequency stability analysis unit is used for dynamically evaluating the frequency change rate of the response signal, and judging whether the liquid column has a frequency jitter behavior under the action of the disturbance pulse signal;
[0040] The attenuation feature extraction unit is used for analyzing the amplitude envelope decay rate of the response signal, and identifying whether the liquid column structure has a non-continuous region or a bubble entrainment behavior based on the joint features of the frequency jitter amplitude and the decay rate.
[0041] Further settings are as follows:
[0042] The state recognition module further comprises an abnormal positioning unit,
[0043] The abnormal positioning unit is connected with a plurality of response acquisition sub-units of the response acquisition module, the plurality of response acquisition sub-units are arranged at different positions of the liquid column structure, and are used for acquiring the response signals generated at the arranged positions after the action of the disturbance pulse signal;
[0044] The abnormal positioning unit estimates the position area of the abnormal response in the liquid column structure based on the propagation time difference or amplitude response difference between the response signals of different collection positions, by comparing the arrival order and strength variation law of each response signal in the spatial propagation path of the disturbance signal.
[0045] Further provided are:
[0046] The state recognition module further comprises an abnormal type recognition unit,
[0047] The abnormal type recognition unit is used for classifying and recognizing the abnormal type of the liquid column structure based on the multi-dimensional feature indicators of the response signal, the multi-dimensional feature indicators comprising a propagation time delay variation rate, a maximum amplitude mutation amount, a frequency jitter standard deviation and an envelope attenuation index, and the abnormal type comprising bubble entrainment, local obstruction, fluid interruption and structure rupture.
[0048] If the propagation time delay variation rate is greater than a first preset threshold value, and the frequency jitter standard deviation exceeds a second preset threshold value, the bubble entrainment is recognized.
[0049] If the envelope attenuation index is greater than a third preset threshold value, and the maximum amplitude mutation amount is less than a fourth preset threshold value, the local obstruction is recognized.
[0050] If there is intermittent interruption in the response signal and the amplitude abnormally rises after recovery, the fluid interruption or structure rupture is recognized.
[0051] Further provided are:
[0052] The abnormal type recognition unit constructs a weighted scoring function based on the multi-dimensional feature indicators, the weighted scoring function being used for calculating the matching score corresponding to each abnormal type, and the abnormal type recognition unit performing optimal matching recognition according to the scoring results of each abnormal type,
[0053] The weighted scoring function assigns preset weights to the propagation time delay variation rate, the maximum amplitude mutation amount, the frequency jitter standard deviation and the envelope attenuation index and performs weighted summation, for multi-factor fusion judgment of the abnormal response features.
[0054] The present application provides a process state monitoring system for new energy battery manufacturing, which is aimed at the liquid column state abnormal problem in the battery manufacturing process, and a monitoring closed-loop system composed of a disturbance excitation module, a response collection module, a state recognition module, a control instruction generation module and a liquid injection execution module is constructed. Through the functional cooperation and signal coupling between the modules, the present application achieves the following comprehensive technical effects in process monitoring, abnormal recognition, spatial positioning and control feedback:
[0055] 1. The disturbance excitation module actively applies a structural disturbance signal to the liquid column, providing an excitation source for non-intrusive detection; the response acquisition module acquires the response signal of the liquid column under disturbance in real time in a synchronous arrangement, breaking the limitation of traditional passive monitoring that can only perceive severe abnormalities;
[0056] The abnormality sensing sensitivity is improved through the disturbance-response coupling mechanism, realizing the cooperation of active sensing and passive response, and greatly improving the response recognition ability of early abnormalities such as bubble entrainment, local blockage, and structure rupture.
[0057] 2. The recognition accuracy is improved through the coupling of feature multi-dimensional extraction and state recognition module. The state recognition module extracts multi-dimensional feature indexes such as propagation delay, amplitude mutation, frequency jitter, and attenuation index; abnormal classification is realized through logical rule judgment, and a weighted scoring function is further constructed to fuse multi-dimensional indexes, improving the discrimination robustness; the primary discrimination of "whether abnormal" is upgraded to high-precision identification of "abnormal type", which is convenient for subsequent targeted adjustment of process strategy.
[0058] 3. The spatial position estimation of abnormal structures (such as bubbles) in the liquid column is realized through the time sequence difference and amplitude propagation model of multi-point response signals; combined with the state recognition result, the system can output "abnormal type + abnormal position" dual information at the same time;
[0059] It is convenient for the control module to perform targeted compensation (such as deceleration, local exhaust, and suspension), reduce invalid intervention actions, and reduce process error risk.
[0060] 4. The control instruction generation module forms executable adjustment instructions according to the abnormality recognition and positioning results;
[0061] The liquid injection execution module dynamically adjusts the liquid injection rate, rhythm or path according to the instructions; the control instruction generation and liquid injection execution modules form a monitoring closed loop, improving the adaptability and stability of the entire process chain. BRIEF DESCRIPTION OF DRAWINGS
[0062] Figure 1 is a system structure schematic diagram of the embodiment. DETAILED DESCRIPTION
[0063] The present application will be further described in detail below in combination with the drawings.
[0064] Embodiment:
[0065] As shown in Figure 1 , a process state monitoring system for new energy battery manufacturing includes:
[0066] A disturbance excitation module is arranged outside the structure of the electrolyte injection path, used to apply a controllable disturbance signal to the liquid column during the injection process;
[0067] The response acquisition module is correspondingly arranged with the disturbance excitation module and is configured to acquire a response signal of the liquid column under the action of the disturbance;
[0068] The state identification module is connected with the response acquisition module and is configured to identify a liquid column structure state based on the response signal and determine whether there is a liquid column discontinuous region or a bubble entrainment behavior;
[0069] The control instruction generation module is connected with the state identification module and is configured to generate a deviation correction control instruction when it is determined that the liquid column state is abnormal.
[0070] The liquid injection execution module is connected with the control instruction generation module and is configured to receive the deviation correction control instruction and perform a liquid injection pause, air exhaust or liquid injection rate adjustment operation.
[0071] In the embodiment, the disturbance excitation module cooperates with the liquid injection execution module to construct a multi-module disturbance control structure including a beat identification logic module, a disturbance pulse generation path and a timing control strategy unit.
[0072] Specifically, the beat identification logic module is configured to receive a periodic signal of the liquid injection execution module in real time and determine an intermittent period in the liquid injection process based on the signal as a safe window for injecting the disturbance signal. The periodic signal can be obtained by time domain analysis on the control signal of the liquid injection valve action, and the identified intermittent stage is, for example, the waiting period between the completion of liquid injection and the start of the next cycle.
[0073] The disturbance pulse generation path is signal-connected with the beat identification logic module, and when it is detected that there is an effective intermittent period, the disturbance pulse signal with an amplitude lower than the main liquid injection pressure is generated by triggering the disturbance pressure source or the multi-way control valve array on the liquid injection path and is transmitted into the liquid column region through the original liquid injection path.
[0074] To further ensure that the disturbance signal does not interfere with the stability of the liquid injection, a timing control strategy unit is provided, which receives the output of the beat identification module, calculates the center time of the intermittent region and sets the action time window, so that the disturbance pulse signal is injected in a center-symmetric manner.
[0075] The excitation strategy ensures the effectiveness and interference safety of the disturbance pulse, and the structure state can be determined by the subsequent response acquisition module to acquire the disturbance response signal of the liquid column.
[0076] To improve the adaptability and precision of the disturbance excitation, a disturbance control model is introduced in the embodiment to model and control the amplitude and period of the disturbance pulse signal according to the physical properties of the electrolyte and the structural features of the liquid injection path.
[0077] Specifically, the system acquires the liquid injection path length L, liquid dynamic viscosity μ and density ρ in the initialization stage, and dynamically generates the control parameters of the disturbance pulse signal through the disturbance control model, so that the excited disturbance can cause identifiable response signals of the liquid column and will not cause liquid surface fluctuation or backflow.
[0078] The disturbance control model can be represented as a pair of functions: an amplitude function A(μ, ρ, L) and a period function T(L), the form of which will be given in the subsequent embodiments. This structure makes the disturbance control parameters dynamically adjustable and structurally dependent, meeting the adaptation needs of various electrolytes and liquid injection structures.
[0079] In this embodiment, the mathematical composition of the disturbance control model is further given to realize the calculable modeling of the amplitude and period of the disturbance pulse signal.
[0080] First, the amplitude A of the disturbance pulse signal is set according to the following exponential decay function:
[0081]
[0082] Where:
[0083] A0: reference disturbance amplitude;
[0084] λ: empirical decay coefficient;
[0085] μ: liquid viscosity;
[0086] ρ: liquid density;
[0087] L: liquid injection path length;
[0088] e: base of natural logarithm.
[0089] Second, the period T of the disturbance pulse signal is defined as:
[0090]
[0091] Where:
[0092] T1, T2, L0: system empirical constant;
[0093] log: natural logarithm function.
[0094] This structure can dynamically correct the disturbance parameters according to different liquid properties and liquid injection path structures, and is particularly suitable for the path differences existing in multiple battery products.
[0095] Through the above model, the system can automatically generate the optimal disturbance signal according to the liquid injection process parameters without manually setting the disturbance strength and frequency, thereby improving the system stability and recognition accuracy.
[0096] In the embodiment, the response acquisition module comprises at least one non-contact vibration sensor, which is arranged outside the liquid injection path and close to the channel wall region where the liquid column is located, and is used to acquire the structural vibration response signal caused after the liquid column is acted on by the disturbance pulse signal.
[0097] The non-contact vibration sensor can be a laser vibration detector or a piezoelectric film array, which has high frequency response capability and is suitable for the medium-high frequency vibration characteristics generated after the liquid column is disturbed and excited. In order to improve the acquisition accuracy, the sensor is adsorbed on the outer wall of the liquid injection channel through a soft base, so as to realize stable measurement without affecting the original sealing performance of the liquid injection path.
[0098] The response acquisition module is connected with the disturbance excitation module. When the beat recognition logic module recognizes the effective liquid injection interval and triggers the disturbance pulse, the disturbance excitation module synchronously sends a trigger signal to the response acquisition module, so that the response acquisition module starts the data acquisition program and completes the full cycle recording of the response signal within the set time window before and after the disturbance.
[0099] After the collected signal is converted into an analog quantity, it enters the signal preprocessing module. The module first performs low-pass filtering and short-time average denoising on the original vibration signal, and then performs amplitude normalization processing to eliminate the amplitude deviation caused by the change of system gain. The normalized signal is used as the input of the subsequent structural response judgment algorithm, which is used to evaluate whether the liquid column has a non-continuous region or a blocking structure.
[0100] In the embodiment, the response acquisition module ensures that the sampling time sequence is highly matched with the excitation process, and effectively avoids the pollution and interference problems caused by the traditional contact type sensing mode through non-contact structure arrangement and online signal processing, which is suitable for various new energy battery liquid injection scenes.
[0101] In the embodiment, the state recognition module comprises a frequency stability analysis unit and a decay characteristic extraction unit, which are used to judge the liquid column structure state from the frequency stability and energy decay dimensions based on the response signal generated by the liquid column under the action of the disturbance pulse signal.
[0102] The frequency stability analysis unit first performs short-time Fourier transform (STFT) processing on the normalized response signal to obtain the frequency-time variation curve f(t). The frequency jitter amplitude is defined as the standard deviation of the main frequency change per unit time:
[0103]
[0104] Wherein:
[0105] f(t i ): the instantaneous main frequency at the i th moment;
[0106] Mean frequency in time period;
[0107] N: number of sampling points;
[0108] σ f : frequency jitter amplitude;
[0109] When σ f exceeds the set threshold σ th , it is determined that the liquid column has a frequency instability phenomenon, and a structural breakpoint or bubble disturbance is suspected to occur.
[0110] The envelope extraction and exponential fitting are used by the decay feature extraction unit to obtain the amplitude decay feature of the response signal. Let the envelope function be E(t), and the exponential decay coefficient α is obtained by least squares fitting:
[0111] E(t) = E0·e -αt
[0112] After taking the logarithm, we have:
[0113] lnE(t) = lnE0-αt
[0114] The decay rate α can be obtained by linear regression. When α is higher than the preset threshold α th , it is determined that the liquid column has abnormal energy dissipation, and the structure is abnormal.
[0115] The state recognition module sets a combination threshold rule:
[0116] When the following conditions are met, output "abnormal" judgment:
[0117] σ f > σth(frequency jitter is too large);
[0118] α > α th (energy decay is too fast).
[0119] The above joint determination mechanism can effectively avoid single indicator misjudgment and is suitable for identifying abnormal vibration behavior caused by liquid column breakpoint or bubble entrainment in complex liquid filling path structure.
[0120] In this embodiment, the response acquisition module includes a plurality of sensors arranged in sequence along different spatial positions of the liquid column structure, numbered S1, S2, …, S n , and the spacing between each sensor is a constant ΔL, used to collect the response signals generated by the liquid column at each arrangement position after the disturbance pulse signal acts.
[0121] The state recognition module comprises an abnormality positioning unit, which is configured to perform joint analysis in time and amplitude dimensions based on the response signals collected by the plurality of sensors, to identify a structural discontinuous region possibly existing in the liquid column and estimate a spatial position segment where the structural discontinuous region is located.
[0122] Suppose that the time when the system triggers the disturbance pulse signal is t0, and the starting time of the response signal collected by each response sensor is t i The disturbance propagation time delay τ i of the ith position is defined as:
[0123] τ i = t i -t0
[0124] When the difference between the propagation time delays of two adjacent sampling points satisfies:
[0125] |τ i+1 -τ i |> τ tresold
[0126] It is indicated that the disturbance propagation path has a mutation in the spatial segment, which is possibly caused by bubble entrainment, structural breakpoints or local liquid discontinuity.
[0127] Wherein, τ tresold is a preset propagation time delay change judgment threshold of the system.
[0128] Meanwhile, the abnormality positioning unit extracts the maximum amplitude A i of each response signal, and when the amplitude difference of adjacent sensors satisfies:
[0129] |A i+1 -A i |> A tresold
[0130] It is judged that the response energy distribution of the liquid column in the spatial segment is abnormal, which is possibly caused by local impedance change, and it is prompted that the segment has structural discontinuity.
[0131] Wherein, A tresold is a judgment threshold of amplitude response mutation.
[0132] When a spatial segment simultaneously satisfies the two conditions of propagation time delay mutation and amplitude response mutation, i.e.:
[0133] |τ i+1 -τ i |> τ tresold and |A i+1 -A i |> A tresold
[0134] Then the abnormal positioning unit outputs the spatial position [Li, Li+1] corresponding to the segment as a structural abnormal area, for triggering path switching, alarm or correction processing in subsequent process steps.
[0135] The above spatial positioning mechanism can effectively deal with the abnormal identification problem under the condition of complex injection path and long-distance distribution of liquid column in the new energy battery manufacturing process, avoid the problem of being unable to locate specific problem paragraphs due to overall response abnormalities, has the advantages of strong real-time performance and high robustness, and is suitable for a plurality of injection precision process state monitoring systems.
[0136] In the embodiment, the state identification module further comprises an abnormal type identification unit, which is used to classify the types of abnormalities existing in the liquid column structure based on the multi-dimensional feature indexes of the response signals generated by the liquid column under the disturbance excitation.
[0137] Specifically, the abnormal type identification unit is provided with a multi-parameter extraction channel, which receives the multi-channel response signal data collected by the response collection module, and calculates the following four types of feature indexes in the local sliding window:
[0138] Propagation time delay change rate (Δτ): the local propagation speed fluctuation is obtained by calculating the difference between the response times of adjacent sensing points and taking the first-order difference.
[0139] Maximum amplitude mutation (ΔA): the difference between the maximum response amplitudes of adjacent collection points is taken as the index for judging the amplitude jump.
[0140] Frequency jitter standard deviation (σ f ): the fluctuation of the main frequency of the response signal is statistically calculated in a certain analysis window, which is used to reflect the stability of the frequency spectrum of the disturbance signal.
[0141] Envelope attenuation index (α): the signal envelope is fitted by an exponential function, and the rate of its attenuation with time is estimated.
[0142] Based on the above features, the system presets the following abnormal type classification rules:
[0143] Bubble entrainment:
[0144] Features: the propagation time delay change rate Δτ increases significantly, and the frequency jitter standard deviation σ f increases.
[0145] Determination condition: Δτ>Δτ1, σ_f>σ1.
[0146] Local blockage:
[0147] Features: the overall response amplitude is weakened, and the envelope attenuation index is increased.
[0148] Determination condition: α>α1 and ΔA<ΔA1.
[0149] Fluid interruption or structural rupture:
[0150] Characteristics: The signal experiences short-term interruptions, and the amplitude increases abnormally after recovery.
[0151] Judgment condition: There is a response interruption interval, and the recovery signal amplitude A_new > A_thresh.
[0152] The judgment thresholds (Δτ1, σ1, α1, ΔA1, A_thresh) are obtained based on the statistical distribution of sample data or set by engineering experience.
[0153] This classification mechanism allows the system to identify notions in the liquid column structure and further determine the type of notion, thus providing a basis for decision-making in subsequent liquid injection strategy optimization (such as defoaming, deceleration, and path switching).
[0154] In this embodiment, to enhance the accuracy and robustness of the anomaly type identification unit, the system constructs a weighted scoring function based on the multidimensional feature indicators to quantify and score various liquid column structural anomalies, and thereby achieves optimal matching identification.
[0155] The anomaly type identification unit constructs a weighted scoring function based on the multi-dimensional feature indicators. This weighted scoring function is used to calculate a matching score for each anomaly type. The anomaly type identification unit then performs optimal matching based on the scoring results for each anomaly type.
[0156] The weighted scoring function assigns preset weights to the propagation delay change rate, the maximum amplitude mutation amount, the frequency jitter standard deviation, and the envelope attenuation index, and performs a weighted summation to determine the multi-factor fusion of abnormal response characteristics.
[0157] The anomaly type identification unit calculates the corresponding score for each of the preset anomaly types and selects the type with the highest score as the identification result of the current liquid column anomaly.
[0158] This method supports feature priority difference expression for different anomaly types, improves robustness to scenarios with blurred feature boundaries, and adapts to the recognition needs under various combinations of liquids and injection rates.
[0159] In this embodiment, the control command generation module is used to generate corresponding control commands based on the liquid column structure state information output by the state recognition module. This module includes a strategy selection unit and a command control unit.
[0160] Among them, the policy selection unit is used for classifying and identifying the state type output by the state identification module, including but not limited to: liquid column structure continuous state, suspected non-continuous state, obvious abnormal state, existence of bubble entrainment state, etc. For each type of state, the system pre-establishes a control strategy mapping table, and the policy selection unit can automatically match the corresponding control strategy number according to the current state type.
[0161] The instruction regulation unit is used for generating specific control parameter instructions according to the selected strategy, and the control parameters include but are not limited to target liquid injection flow rate, liquid injection duration, disturbance excitation parameters (such as frequency, amplitude), path selection identifier, etc. The control parameter instructions are output to the lower bit liquid injection execution module in the format of standard communication protocol.
[0162] For example, if the state identification module identifies that the liquid column structure has a bubble entrainment behavior, the control instruction generation module will select the "slow injection-disturbance enhancement-path switching" strategy, and generate the following control instruction fields: the liquid injection flow rate is reduced to 50% of the rated value, the disturbance frequency is increased by 20%, and the liquid injection path is switched to the standby channel. The above instructions are parsed and executed by the liquid injection execution module in real time.
[0163] In the embodiment, the liquid injection execution module is used for receiving and executing the control parameter instructions output by the control instruction generation module, and completing the dynamic adjustment of the liquid injection process accordingly.
[0164] The liquid injection execution module includes a flow control unit, a path switching unit and a disturbance synchronization unit:
[0165] The flow control unit receives the target liquid injection flow rate and liquid injection time length parameters, and drives the metering pump or proportional valve execution element to complete the precise liquid delivery;
[0166] The path switching unit controls the liquid injection channel switching valve according to the path switching instruction, and switches to the preset standby path or bypass liquid discharge channel;
[0167] The disturbance synchronization unit is used for synchronously executing the real-time adjustment of the disturbance excitation module parameters, so as to ensure that the optimal disturbance frequency or mode is matched under a specific liquid injection state.
[0168] For example, when the control parameter instruction requires that the liquid injection path is switched from the main channel A to the standby channel B, the liquid flow rate is reduced to 30 mL per minute, and the disturbance pulse frequency is increased to 80 Hz, the liquid injection execution module will complete the three adjustment operations in parallel after receiving the instruction, to ensure that the liquid injection path, rate and disturbance excitation are linked and consistent.
[0169] Through the interface protocol, high-speed data exchange with the control instruction generation module is realized, and the current execution state is also supported to be fed back to the system main control unit, to improve the real-time control and fault response ability of the system.
[0170] The disturbance response recognition result and the control logic can be linked in a closed loop through the liquid injection execution module, thereby ensuring the stability and adaptability of the liquid injection process.
[0171] To sum up, the process state monitoring system for new energy battery manufacturing provided by the application realizes linkage and cooperation of the disturbance excitation module, the response acquisition module, the state recognition module, the control instruction generation module and the liquid injection execution module, and builds a process process active monitoring mechanism based on disturbance-response.
[0172] In practical application, the disturbance excitation module applies a disturbance pulse signal to the target liquid injection area through structural arrangement to stimulate the response behavior of the liquid column structure to the disturbance; the response acquisition module is correspondingly arranged to acquire the multi-dimensional response signals of the liquid column under the action of the disturbance in real time, including propagation time delay, signal amplitude, main frequency jitter and envelope attenuation.
[0173] The state recognition module further extracts a plurality of feature indexes from the acquired response signals, and combines a preset discrimination rule or a weighted scoring function to identify whether there is a non-continuous structure state in the liquid column, and to determine the abnormal type, such as bubble entrainment, local blockage, fluid interruption, etc.
[0174] The control instruction generation module automatically generates liquid injection parameter adjustment instructions according to the recognition result, controls the liquid injection execution module to perform dynamic adjustment operation, and realizes self-adaptive control of the liquid injection process. Through the linkage and cooperation of the above functional modules, the application can realize early identification, type classification and position perception of abnormal state of key liquid injection link in new energy battery manufacturing process, and further realize closed loop feedback control, has the advantages of strong non-invasiveness, high adaptability and good universality, is suitable for various battery liquids, structures and liquid injection process paths, and significantly improves the stability and consistency of new energy battery manufacturing process.
[0175] The above-described embodiments do not constitute a limitation on the protection scope of the technical solutions. Any modifications, equivalent replacements and improvements made within the spirit and principles of the above-described embodiments shall be included in the protection scope of the technical solutions.
Claims
1. A process state monitoring system for new energy battery manufacturing, characterized in that, The method comprises the following steps: A disturbance excitation module is arranged outside the structure of the electrolyte injection path, and is used to apply a controllable disturbance signal to the liquid column during the injection process; A response acquisition module is arranged corresponding to the disturbance excitation module, and is used to acquire the response signal of the liquid column under the action of the disturbance; A state identification module is connected with the response acquisition module, and is used to identify the state of the liquid column based on the response signal, and to determine whether there is a non-continuous region of the liquid column or a bubble entrainment behavior; A control instruction generation module is connected with the state identification module, and is used to generate a correction control instruction when the state of the liquid column is abnormal; An injection execution module is connected with the control instruction generation module, and is used to receive the correction control instruction and execute the injection pause, exhaust or injection rate adjustment operation.
2. The process state monitoring system for new energy battery manufacturing according to claim 1, characterized in that, The disturbance excitation module comprises: A beat identification logic module is used to receive a control cycle signal from the injection execution module, and identify the intermittent stage in the injection cycle based on the signal; A disturbance pulse generation path is connected with the beat identification logic module, and is used to receive a trigger signal in the intermittent stage, and control a disturbance pressure source or an injection control valve array to generate a disturbance pulse signal with an amplitude lower than the main injection pressure; A timing control strategy unit is connected with the beat identification logic module and the disturbance pulse generation path, and is used to set the trigger time point and action time of the disturbance pulse signal in the intermittent stage, so that the disturbance pulse signal is applied to the liquid column in a center-symmetric manner through the injection channel, and is used to excite the response signal of the liquid column under the action of the disturbance.
3. The process state monitoring system for new energy battery manufacturing according to claim 2, characterized in that, The disturbance pulse generation path determines the amplitude and period of the disturbance pulse signal based on a disturbance control model, which is constructed according to the physical properties of the liquid column or the structure parameters of the injection path, so that the generated disturbance pulse signal has dynamic adaptive characteristics sufficient to excite the response signal of the liquid column and does not interfere with the injection process.
4. The process state monitoring system for new energy battery manufacturing according to claim 3, characterized in that, The disturbance control model comprises an amplitude function and a period function, The amplitude function A adopts an exponential decay form, and is constructed based on the dynamic viscosity μ, density ρ and injection path length L of the liquid as follows: The period function T adopts a segmented response form, and is adjusted based on the path length L and the natural frequency interval of the liquid column, and its expression is as follows: Wherein, A0 is the reference disturbance amplitude; λ is the empirical decay coefficient; T1, T2 and L0 are empirical constant; e is the base of natural logarithm; 1og is the natural logarithm function.
5. The process state monitoring system for new energy battery manufacturing according to claim 1, characterized in that, The response acquisition module comprises at least one non-contact vibration sensor, which is used to acquire the vibration response signal generated after the disturbance pulse signal acts on the liquid column; the response acquisition module realizes synchronous acquisition by receiving the trigger signal from the disturbance excitation module, and performs preset signal denoising and amplitude normalization processing on the acquired vibration response signal.
6. The process state monitoring system for new energy battery manufacturing according to claim 5, characterized in that, The state identification module comprises a frequency stability analysis unit and a decay characteristic extraction unit, The frequency stability analysis unit is used to dynamically evaluate the frequency change rate of the response signal, and determine whether there is a frequency jitter behavior of the liquid column under the action of the disturbance pulse signal; The attenuation feature extraction unit is configured to analyze the amplitude envelope decay rate of the response signal, and identify whether the liquid column structure has a non-continuous region or bubble entrainment behavior based on a joint feature of the frequency jitter amplitude and the decay rate.
7. The process state monitoring system for new energy battery manufacturing according to claim 6, characterized in that, The state recognition module further comprises an anomaly positioning unit, The anomaly positioning unit is connected with a plurality of response acquisition sub-units of the response acquisition module, the plurality of response acquisition sub-units are arranged at different positions of the liquid column structure, and are configured to acquire response signals generated at the arranged positions after the disturbance pulse signal acts; The anomaly positioning unit estimates the position region of the response anomaly in the liquid column structure based on the propagation time difference or amplitude response difference between the response signals at different acquisition positions, by comparing the arrival order and strength variation law of each response signal in the spatial propagation path of the disturbance signal.
8. The process state monitoring system for new energy battery manufacturing according to claim 7, characterized in that, The state recognition module further comprises an anomaly type recognition unit, The anomaly type recognition unit is configured to classify and identify the anomaly type of the liquid column structure based on a plurality of dimensional feature indexes of the response signal, the dimensional feature indexes comprising a propagation time delay variation rate, a maximum amplitude mutation amount, a frequency jitter standard deviation, and an envelope decay index, and the anomaly type comprising bubble entrainment, local obstruction, fluid interruption, and structure rupture; If the propagation time delay variation rate is greater than a first preset threshold value, and the frequency jitter standard deviation is greater than a second preset threshold value, the anomaly type is identified as bubble entrainment; If the envelope decay index is greater than a third preset threshold value, and the maximum amplitude mutation amount is less than a fourth preset threshold value, the anomaly type is identified as local obstruction; If there is intermittent interruption in the response signal and the amplitude abnormally rises after recovery, the anomaly type is identified as fluid interruption or structure rupture.
9. The process state monitoring system for new energy battery manufacturing according to claim 8, characterized in that, The anomaly type recognition unit constructs a weighted scoring function based on the plurality of dimensional feature indexes, the weighted scoring function being used to calculate a matching score corresponding to each anomaly type, and the anomaly type recognition unit performing optimal matching identification according to the scoring results of each anomaly type, The weighted scoring function assigns preset weights to the propagation time delay variation rate, the maximum amplitude mutation amount, the frequency jitter standard deviation, and the envelope decay index, and performs weighted summation, for multi-factor fusion judgment of the anomaly response feature.