From: azao <821478748@qq.com> Date: Wed, 29 Oct 2025 10:21:54 +0000 (+0800) Subject: 柒牌： X-Git-Url: http://git.snicker.com.cn/?a=commitdiff_plain;h=6ac0882ad1613959fed210306ebfa298c4ac0568;p=mirrors%2FASR1603_G5.git 柒牌： 添加语音更改nlp判断 完善AI翻译功能 金三： 修改语音对话中断的bug --- diff --git a/2.sdk/project/7PIN_1_5/sdk/services/microPython/ports/quectel/core/ai_VC.py b/2.sdk/project/7PIN_1_5/sdk/services/microPython/ports/quectel/core/ai_VC.py index 0ad3f43f..5f5e0e85 100644 --- a/2.sdk/project/7PIN_1_5/sdk/services/microPython/ports/quectel/core/ai_VC.py +++ b/2.sdk/project/7PIN_1_5/sdk/services/microPython/ports/quectel/core/ai_VC.py @@ -2511,11 +2511,13 @@ class AI_VC(Abstract): self.play_new_voice_flag =0 #播放文字留言flag EventMesh.subscribe("play_tts",self.play_tts) EventMesh.subscribe("play_new_voice",self.play_new_voice) - EventMesh.subscribe("play_music",self.play_music) + EventMesh.subscribe("play_music",self.play_music)#播放音乐 + EventMesh.subscribe("play_trans_tts",self.play_trans_tts)#接收翻译tts EventMesh.subscribe("set_vad_silence",self.set_vad_silence) EventMesh.subscribe("kws_even",self.kws_even) EventMesh.subscribe("change_volume",self.change_volume) - EventMesh.subscribe("find_cmd",self.find_cmd) + EventMesh.subscribe("find_cmd",self.find_cmd) #查询命令 + EventMesh.subscribe("change_voice",self.change_voice) #改变音色 self.voice_timer = osTimer()#语音留言10S按下按键确定是否播放的判断定时器 self.voice_cnt = 0 #留言定时器计数 self.voicerupt_flag = 0#留言选择按键接听标志位 @@ -4806,32 +4808,32 @@ class AI_VC(Abstract): self.AI_TTS_str ="服务器地址"+self.server_ip return 0 - if text.find("讲话") != -1 or text.find("说话") != -1: - if text.find("å¿«") != -1 and text.find("太") == -1 : - self.speech_rate += 10 - if self.speech_rate > 50: - - self.AI_TTS_str = "语速已经是最快了" - self.speech_rate = 50 - else: + # if text.find("讲话") != -1 or text.find("说话") != -1: + # if text.find("å¿«") != -1 and text.find("太") == -1 : + # self.speech_rate += 10 + # if self.speech_rate > 50: + + # self.AI_TTS_str = "语速已经是最快了" + # self.speech_rate = 50 + # else: - EventMesh.publish("set_storage_config",["speech_rate",self.speech_rate]) + # EventMesh.publish("set_storage_config",["speech_rate",self.speech_rate]) - self.AI_TTS_str = "明白，语速已经调快了,当前语速是{}".format(self.speech_rate) + # self.AI_TTS_str = "明白，语速已经调快了,当前语速是{}".format(self.speech_rate) - return 0 - if text.find("慢") != -1 and text.find("太") == -1 : - self.speech_rate -= 10 - if self.speech_rate < -50: - self.speech_rate = -50 - #EventMesh.publish("audio_montage_play",["yinliangyizuixiao.mp3"]) - #self.aud.play(1,0,'U:/static/yinliangyizuixiao.mp3')#音量已最小 - self.AI_TTS_str = "语速已经是最慢了" - else: - EventMesh.publish("set_storage_config",["speech_rate",self.speech_rate]) - - self.AI_TTS_str = "明白，语速已经调慢了,当前语速是{}".format(self.speech_rate) - return 0 + # return 0 + # if text.find("慢") != -1 and text.find("太") == -1 : + # self.speech_rate -= 10 + # if self.speech_rate < -50: + # self.speech_rate = -50 + # #EventMesh.publish("audio_montage_play",["yinliangyizuixiao.mp3"]) + # #self.aud.play(1,0,'U:/static/yinliangyizuixiao.mp3')#音量已最小 + # self.AI_TTS_str = "语速已经是最慢了" + # else: + # EventMesh.publish("set_storage_config",["speech_rate",self.speech_rate]) + + # self.AI_TTS_str = "明白，语速已经调慢了,当前语速是{}".format(self.speech_rate) + # return 0 if len(text)<22 and(text.find("聆听") != -1 or text.find("明天")!= -1) : #print("text len :{}".format(len(text))) if text.find("å¿«") != -1 and text.find("太") == -1 : @@ -5490,27 +5492,50 @@ class AI_VC(Abstract): battery = EventMesh.publish("get_battery") print("电量查询：{}".format(battery)) tts_play_str = "电量剩余百分之{}".format(battery) - EventMesh.publish("play_tts",[tts_play_str,1]) - elif "信号" in msg: - signal = net.csqQueryPoll() - tts_play_str = "当前信号{}".format(signal) - EventMesh.publish("play_tts",[tts_play_str,1]) - elif "卡号" in msg: - iccid = sim.getIccid() - if iccid != -1 and len(iccid) > 5: - print("iccid{}".format(iccid)) - tts_play_str = "当前卡号为:{}".format(iccid) - else: - tts_play_str = "卡号查询失败" - EventMesh.publish("play_tts",[tts_play_str,1]) - elif "设备号" in msg: - imei = modem.getDevImei() - if imei != -1 and len(imei) > 5: - print("IMEI：{}".format(imei)) - tts_play_str = "当前设备号为:{}".format(imei) - else: - tts_play_str = "设备号查询失败" - EventMesh.publish("play_tts",[tts_play_str,1]) + EventMesh.publish("play_tts",[tts_play_str,0]) + #改变音色判断 + def change_voice(self,event ,msg): + voice = msg + if "男声" in msg or "普通话" in msg: + self.VC_TTS_voice_key = "男" + elif "女声" in msg: + self.VC_TTS_voice_key = "女" + elif "童声" in msg: + self.VC_TTS_voice_key = "ç«¥" + elif "香港粤语" in msg: + self.VC_TTS_voice_key = "香港粤语" + elif "四川话" in msg: + self.VC_TTS_voice_key = "四川" + elif "上海话" in msg: + self.VC_TTS_voice_key = "上海" + elif "安徽合肥话" in msg: + self.VC_TTS_voice_key = "安徽合肥" + elif "山东话" in msg: + self.VC_TTS_voice_key = "山东话" + elif "陕西话" in msg: + self.VC_TTS_voice_key = "陕西" + elif "广东话" in msg: + self.VC_TTS_voice_key = "广东" + elif "成都话" in msg: + self.VC_TTS_voice_key = "成都" + elif "湖南话" in msg: + self.VC_TTS_voice_key = "湖南" + elif "苏州话" in msg: + self.VC_TTS_voice_key = "苏州" + elif "东北话" in msg: + self.VC_TTS_voice_key = "东北" + elif "河南话" in msg: + self.VC_TTS_voice_key = "河南" + elif "湖北话" in msg: + self.VC_TTS_voice_key = "湖北" + elif "闽南话" in msg: + self.VC_TTS_voice_key = "闽南" + elif "内蒙古话" in msg: + self.VC_TTS_voice_key = "内蒙古话" + EventMesh.publish("set_storage_config",["VC_TTS_voice_key",self.VC_TTS_voice_key]) + tts_play_str = "声音改为{}".format(voice) + EventMesh.publish("play_tts",[tts_play_str,0]) + #语音留言 def play_new_voice(self,event = None,msg= None): EventMesh.publish("publish_AI_btnInfo",{"func":self.play_new_voice_cb,"data":msg}) @@ -5571,6 +5596,31 @@ class AI_VC(Abstract): # EventMesh.publish("publish_AI_btnInfo",{"func":self.AI_VC_start}) except Exception as e: usys.print_exception(e) + + #AI翻译接收TTS + def play_trans_tts(self,event = None,msg= None): + EventMesh.publish("publish_AI_btnInfo",{"func":self.play_trans_tts_cb,"data":msg}) + def play_trans_tts_cb(self,data): + try: + new_voice_url = data #音频地址 + if new_voice_url != "": + self.AI_new_voice_url = new_voice_url + if self.AI_VC_send_data_queue.empty()!=True: + self.AI_VC_send_data_queue.clear() + if self.AI_ASR_send_data_queue.empty()!=True: + self.AI_ASR_send_data_queue.clear() + EventMesh.publish("open_led_all") + self.AI_ASR_send_buf_len = 0 + self.VC_Speech_End() + self.websocket_deconnect_VC() + self.AI_VC_start() + EventMesh.publish("publish_AI_btnInfo",{"func":self.AI_client_new_voice})#播放留言 + self.play_new_voice_flag = 1 + # EventMesh.publish("publish_AI_btnInfo",{"func":self.VC_Speech_End}) + # EventMesh.publish("publish_AI_btnInfo",{"func":self.websocket_deconnect_VC}) + # EventMesh.publish("publish_AI_btnInfo",{"func":self.AI_VC_start}) + except Exception as e: + usys.print_exception(e) def AI_client_new_voice(self): #上传未读的语音地址/包括音频地址也可以 json_data = \ { diff --git a/2.sdk/project/7PIN_1_5/sdk/services/microPython/ports/quectel/core/main_init.py b/2.sdk/project/7PIN_1_5/sdk/services/microPython/ports/quectel/core/main_init.py index 460ac182..68ab7349 100644 --- a/2.sdk/project/7PIN_1_5/sdk/services/microPython/ports/quectel/core/main_init.py +++ b/2.sdk/project/7PIN_1_5/sdk/services/microPython/ports/quectel/core/main_init.py @@ -250,6 +250,7 @@ def main_init(): pcr.add_support(pro.IOTWeb_user_unbind_devProtoCol()) pcr.add_support(pro.IOTWeb_play_MusicProtoCol()) pcr.add_support(pro.IOTWeb_Nlp_ConfigProtoCol()) + pcr.add_support(pro.IOTWeb_play_trans_TTSProtoCol()) sm = pro.SockMangager() sm.add_support_resolver(pcr) diff --git a/2.sdk/project/7PIN_1_5/sdk/services/microPython/ports/quectel/core/protocol.py b/2.sdk/project/7PIN_1_5/sdk/services/microPython/ports/quectel/core/protocol.py index 0b109eaf..75a413b0 100644 --- a/2.sdk/project/7PIN_1_5/sdk/services/microPython/ports/quectel/core/protocol.py +++ b/2.sdk/project/7PIN_1_5/sdk/services/microPython/ports/quectel/core/protocol.py @@ -2597,12 +2597,21 @@ def ai_cmd_handle(cmd): find = cmd[voice_start_index:]#查询截取 if "å¿«" in find: EventMesh.publish("rate_move",1) + volume = "变快" elif "慢" in find: EventMesh.publish("rate_move",0) + volume = "变慢" + tts_play_str = "语速已经" + volume + EventMesh.publish("play_tts",[tts_play_str,0]) elif "会议纪要" in cmd:#当指令中有会议纪要时，进入会议纪要模式 EventMesh.publish("come_meet_mode")#进入会议纪要 elif "翻译" in cmd:#当指令中有翻译时，进入AI翻译模式 EventMesh.publish("come_trans_mode")#进入AI翻译纪要 + elif "声音改为" in cmd:#当指令中有声音改为时，进入音色修改 + voice_start_index = cmd.index("声音改为") + len("声音改为")#动态判断截取部分 + find = cmd[voice_start_index:]#查询截取 + EventMesh.publish("change_voice",find) + #NLP指令 class IOTWeb_Nlp_ConfigProtoCol(ProtoCol): """ai_cmd""" @@ -2635,6 +2644,38 @@ class IOTWeb_Nlp_ConfigProtoCol(ProtoCol): ai_cmd_handle(cmdName) print("recv push_music{}".format(msg)) +#翻译播放TTS +class IOTWeb_play_trans_TTSProtoCol(ProtoCol): + """tts_audio_url""" + TYPE = "tts_audio_url" + + def send(self, event, msg): + # print("tts_audio_url{}".format(msg)) + # 通过websocket发送到服务器端 + date_tmp = utime.localtime() + time_tmp = utime.mktime(date_tmp) + tts_audio_url = { + "cmd":"tts_audio_url_complete", + "ts":time_tmp, + "msg":"返回状态", + "data":None + } + send_data = b'' + send_data = protocol_send_total_data(tts_audio_url) + print("send_data tts_audio_url{}".format(send_data)) + EventMesh.publish("CHANNEL_IOT_WEB_SEND", send_data) + + def recv(self, event, msg): + """ + @param event:上行下发push_music + @param msg: 接收到的下行数据 + """ + #提取录音播放名称 + voiceUrl = msg["data"] + print("提取到的翻译路径为{}".format(voiceUrl)) + EventMesh.publish("play_trans_tts",voiceUrl) + print("recv tts_audio_url{}".format(msg)) + #音乐播放 class IOTWeb_play_MusicProtoCol(ProtoCol): """push_music""" diff --git a/2.sdk/project/K8/sdk/services/microPython/ports/quectel/core/ai_VC.py b/2.sdk/project/K8/sdk/services/microPython/ports/quectel/core/ai_VC.py index 837e970f..1324e7b0 100644 --- a/2.sdk/project/K8/sdk/services/microPython/ports/quectel/core/ai_VC.py +++ b/2.sdk/project/K8/sdk/services/microPython/ports/quectel/core/ai_VC.py @@ -1372,6 +1372,7 @@ class AI_VC(Abstract): self.Close_flag =0 #=1时，连续对话播放完音频就退出 self.AI_alarm = _ai_alarm.AI_alarm() self.buffer = bytearray(1024*8) #音频缓存区 + self.tts_completed = 0#音频是否播放过 self.stop_tts = 0#音频是否播放完 #表情变化状态以及检索词 self.face_status = 0 # 0为正常，1为哭泣，2为愤怒，3为疑惑 @@ -3149,6 +3150,7 @@ class AI_VC(Abstract): elif jsload_data["event"] == "tts.completed" :#tts结束 EventMesh.publish("open_led_green")#绿色 EventMesh.publish("publish_AI_btnInfo",{"func":self.VC_Play_Start}) + self.tts_completed = 1 self.LocalRespondingEnded_timerout = 0 self.VC_state = VC_State_TTSEnd pass @@ -3163,9 +3165,10 @@ class AI_VC(Abstract): #EventMesh.publish("open_led_green")#绿色 self.LocalRespondingEnded_timerout = 3 self.VC_state = VC_State_End - if self.stop_tts == 1: + if self.stop_tts == 1 and self.tts_completed == 1:#已经播放了音频并且音频停止了 self.AI_client_VC_end_tts()#音频结束 self.stop_tts = 0 + self.tts_completed = 0 EventMesh.publish("send_device_VC_state")#上传VC状态 pass elif jsload_data["event"] == "command" :#命令匹配